Teach Yourself UNIX

Transcription

Teach Yourself UNIX

Teach Yourself
UNIX
in 24 Hours
Dave Taylor
James C. Armstrong, Jr.
201 West 103rd Street
Indianapolis, Indiana 46290
Teach Yourself UNIX in 24 Hours
iv
Decimilli accipitrae Raptor Regina.—JA
To the newest light of my life: Ashley Elizabeth.—DT
Copyright  1997 by Sams Publishing
FIRST EDITION
All rights reserved. No part of this book shall be reproduced, stored in a
retrieval system, or transmitted by any means, electronic, mechanical, photocopying, recording, or otherwise, without written permission from the
publisher. No patent liability is assumed with respect to the use of the information contained herein. Although every precaution has been taken in the
preparation of this book, the publisher and author assume no responsibility for
errors or omissions. Neither is any liability assumed for damages resulting from
the use of the information contained herein. For information, address Sams
Publishing, 201 W. 103rd St., Indianapolis, IN 46290.
International Standard Book Number: 0-672-31107-0
Library of Congress Catalog Card Number: 97-66198
2000
99
98
97
4
3
2
1
Acquisitions Editor
Grace M. Buechlein
Development Editor
Brian-Kent Proffitt
Production Editor
Kristi Hart
Indexer
Greg Pearson
Technical Reviewer
Raj Mangal
Editorial Coordinators
Mandi Rouell
Katie Wise
Technical Edit Coordinator
Lynette Quinn
Resource Coordinator
Deborah Frisby
Interpretation of the printing code: the rightmost double-digit number is the
year of the book’s printing; the rightmost single-digit, the number of the book’s
printing. For example, a printing code of 97-1 shows that the first printing of
the book occurred in 1997.
Editorial Assistants
Carol Ackerman
Andi Richter
Rhonda Tinch-Mize
Composed in AGaramond and MCPdigital by Macmillan Computer Publishing
Cover Designer
Tim Amrhein
Printed in the United States of America
All terms mentioned in this book that are known to be trademarks or service
marks have been appropriately capitalized. Sams Publishing cannot attest to the
accuracy of this information. Use of a term in this book should not be regarded
as affecting the validity of any trademark or service mark.
Book Designer
Gary Adair
Copy Writer
David Reichwein
Production Team
Supervisors
Brad Chinn
Charlotte Clapp
President, Sams Publishng Richard K. Swadley
Publishing Manager Dean Miller
Director of Editorial Services Cindy Morrow
Director of Marketing Kelli Spencer
Product Marketing Manager Wendy Gilbride
Assistant Marketing Managers Jen Pock, Rachel Wolfe
Production
Brad Lenser
Chris Livengood
Gene Redding
Janet Seib
Overview
Hour 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Introduction
What Is this UNIX Stuff?
Getting onto the System and Using the Command Line
Moving About the File System
Listing Files and Managing Disk Usage
Ownership and Permissions
Creating, Moving, Renaming, and Deleting Files and Directories
Looking into Files
Filters and Piping
Wildcards and Regular Expressions
Power Filters and File Redirection
An Introduction to the vi Editor
Advanced vi Tricks, Tools, and Techniques
An Overview of the emacs Editor
Introduction to Command Shells
Getting the Most Out of the C Shell
Basic Shell Programming
Job Control
Printing in the UNIX Environment
Searching for Information and Files
Communicating with Others
Using Netscape To See the World Wide Web
Internet E-Mail, Netnews, and IRC
Using telnet and ftp
Programming in C for UNIX
Glossary
xvi
1
21
43
63
87
113
127
145
161
187
199
245
281
305
323
347
361
379
397
407
425
443
479
509
531
Index
541
vi
Contents
Hour 1
What Is This UNIX Stuff?
1
Goals for This Hour ................................................................................... 1
What Is UNIX? .......................................................................................... 2
A Brief History of UNIX ............................................................................ 3
The C Programming Language .............................................................. 4
UNIX Becomes Popular ........................................................................ 5
What’s All This About Multiuser Systems? ................................................. 5
Cracking Open the Shell ............................................................................ 6
Getting Help .............................................................................................. 7
Task 1.1: Man Pages, UNIX Online Reference ...................................... 7
Task 1.2: Other Ways to Find Help in UNIX ..................................... 14
Summary .................................................................................................. 17
Workshop ................................................................................................ 17
Key Terms ........................................................................................... 17
Questions ............................................................................................ 18
Preview of the Next Hour .................................................................... 19
2
21
Goals for This Hour ................................................................................. 21
Task 2.1: Logging In and Out of the System ....................................... 22
Task 2.2: Changing Passwords with passwd ........................................ 25
Task 2.3: Picking a Secure Password .................................................... 26
Task 2.4: Who Are You? ...................................................................... 28
Task 2.5: Finding Out What Other Users Are Logged
in to the System ................................................................................. 30
Task 2.6: What Is Everyone Doing on the Computer? ......................... 31
Task 2.7: Checking the Current Date and Time .................................. 33
Task 2.8: Looking at a Calendar .......................................................... 33
Simple Math with UNIX .......................................................................... 36
Task 2.9: Using the bc Infix Calculator ............................................... 36
Task 2.10: Using the dc Postfix Calculator .......................................... 38
Summary .................................................................................................. 40
Workshop ................................................................................................ 40
Key Terms ........................................................................................... 40
Questions ............................................................................................ 41
3
Moving About the File System
43
What a Hierarchical File System Is All About ........................................... 44
Task 3.1: The UNIX File System Organization ................................... 45
The bin Directory ............................................................................... 46
The dev Directory ............................................................................... 47
The etc Directory ............................................................................... 47
The lib Directory ............................................................................... 47
The lost+found Directory .................................................................. 48
The mnt and sys Directories ............................................................... 48
The tmp Directory ............................................................................... 48
The usr Directory ............................................................................... 48
Other Miscellaneous Stuff at the Top Level ......................................... 49
How Mac and PC File Systems Differ from the UNIX File System .......... 50
Directory Separator Characters ................................................................. 50
The Difference Between Relative and Absolute Filenames ........................ 51
Task 3.2: Hidden Files in UNIX ......................................................... 52
Task 3.3: The Special Directories “.” and “..” ...................................... 55
Task 3.4: The env Command .............................................................. 56
Task 3.5: PATH and HOME ..................................................................... 57
Task 3.6: Find Where You Are with pwd ............................................. 58
Task 3.7: Move to Another Location with cd ...................................... 58
Summary .................................................................................................. 60
Workshop ................................................................................................ 60
Key Terms ........................................................................................... 60
Questions ............................................................................................ 62
4
63
The ls Command ............................................................................... 64
Task 4.1: All About the ls Command ................................................. 64
Task 4.2: Having ls Tell You More .................................................... 65
Task 4.3: Combining Flags .................................................................. 68
Task 4.4: Listing Directories Without Changing Location ................... 69
Special ls Command Flags ....................................................................... 71
Task 4.5: Changing the Sort Order in ls ............................................. 71
Task 4.6: Listing Directory Trees Recursively in ls ............................. 73
Task 4.7: Long Listing Format in ls ................................................... 74
Permissions Strings .............................................................................. 74
Task 4.8: Long Listing Format for Directories in ls ............................ 75
Task 4.9: Creating Files with the touch Command ............................. 78
Task 4.10: Check Disk-Space Usage with du ....................................... 79
Task 4.11: Check Available Disk Space with df ................................... 82
Task 4.12: Shrink Big Files with the compress Program ..................... 83
Summary .................................................................................................. 84
Workshop ................................................................................................ 84
Key Terms ........................................................................................... 84
Questions ............................................................................................ 85
viii
5
87
Task 5.1: Understand File Permissions Settings ................................... 88
Task 5.2: Directory Permissions Settings ............................................. 93
Task 5.3: Modify File and Directory Permissions with chmod .............. 96
Task 5.4: Set New File Permissions with chmod ................................... 98
Task 5.5: Calculating Numeric Permissions Strings ........................... 102
Task 5.6: Establish Default File and Directory Permissions
with the umask Command .............................................................. 104
Task 5.7: Identify Owner and Group for Any File or Directory ......... 107
Task 5.8: Change the Owner of a File or Directory ........................... 108
Task 5.9: Change the Group of a File or Directory ............................ 109
Summary ................................................................................................ 110
Workshop .............................................................................................. 110
Key Terms ......................................................................................... 110
Questions .......................................................................................... 111
Preview of the Next Hour .................................................................. 111
6
7
Creating, Moving, Renaming, and Deleting Files
and Directories
113
Goals for This Hour ............................................................................... 113
Task 6.1: Creating New Directories Using mkdir .............................. 114
Task 6.2: Copying Files to New Locations Using cp .......................... 116
Task 6.3: Moving Files to New Locations Using mv ........................... 118
Task 6.4: Renaming Files with mv ...................................................... 119
Task 6.5: Removing Directories with rmdir ...................................... 120
Task 6.6: Removing Files Using rm .................................................... 121
Task 6.7: Minimizing the Danger of the rm Command ..................... 123
Summary ................................................................................................ 125
Workshop .............................................................................................. 125
Key Terms ......................................................................................... 125
Questions .......................................................................................... 126
Looking into Files
127
Task 7.1: Using file to Identify File Types ...................................... 128
Task 7.2: Exploring UNIX Directories with file .............................. 130
Task 7.3: Peeking at the First Few Lines with head ........................... 133
Task 7.4: Viewing the Last Few Lines with tail ............................... 135
Task 7.5: Viewing the Contents of Files with cat .............................. 136
Task 7.6: Viewing Larger Files with more .......................................... 139
Summary ................................................................................................ 143
Workshop .............................................................................................. 143
Key Terms ......................................................................................... 143
Questions .......................................................................................... 144
Contents
8
Filters and Piping
ix
145
Task 8.1: The Secrets of File Redirection ........................................... 146
Task 8.2: Counting Words and Lines Using wc ................................. 147
Task 8.3: Removing Extraneous Lines Using uniq ............................ 149
Task 8.4: Sorting Information in a File Using sort ........................... 150
Task 8.5: Number Lines in Files Using cat -n and nl ....................... 153
Task 8.6: Cool nl Tricks and Capabilities ......................................... 154
Summary ................................................................................................ 157
Workshop .............................................................................................. 158
Key Terms ......................................................................................... 158
Questions .......................................................................................... 158
9
10
11
161
Task 9.1: Filename Wildcards ............................................................ 162
Task 9.2: Advanced Filename Wildcards ........................................... 164
Task 9.3: Creating Sophisticated Regular Expressions ........................ 167
Task 9.4: Searching Files Using grep ................................................. 172
Task 9.5: For Complex Expressions, Try egrep ................................. 175
Task 9.6: Searching for Multiple Patterns at Once with fgrep .......... 176
Task 9.7: Changing Things En Route with sed ................................. 179
Summary ................................................................................................ 185
Workshop .............................................................................................. 185
Key Terms ......................................................................................... 185
Questions .......................................................................................... 185
187
Task 10.1: The Wild and Weird awk Command ............................... 188
Task 10.2: Re-routing the Pipeline with tee ...................................... 196
Summary ................................................................................................ 197
Workshop .............................................................................................. 197
Questions .......................................................................................... 197
199
Task 11.1: How To Start and Quit vi ............................................... 200
Task 11.2: Simple Cursor Motion in vi ............................................ 205
Task 11.3: Moving by Words and Pages ............................................ 208
Task 11.4: Inserting Text into the File Using i, a, o, and O ............... 212
Task 11.5: Deleting Text ................................................................... 220
Task 11.6: Searching Within a File .................................................... 229
Task 11.7: How To Start vi Correctly .............................................. 234
Task 11.8: The Colon Commands in vi ........................................... 236
x
Summary ................................................................................................ 242
Workshop .............................................................................................. 243
Key Terms ......................................................................................... 243
Questions .......................................................................................... 244
12
13
14
245
Task 12.1: The Change and Replace Commands ............................... 246
Task 12.2: Numeric Repeat Prefixes .................................................. 253
Task 12.3: Numbering Lines in the File ............................................ 255
Task 12.4: Search and Replace ........................................................... 257
Task 12.5: Mapping Keys with the :map Command .......................... 260
Task 12.6: Moving Sentences and Paragraphs .................................... 266
Task 12.7: Access UNIX with ! ......................................................... 270
Summary of vi Commands .................................................................... 278
Summary ................................................................................................ 279
Workshop .............................................................................................. 279
Key Terms ......................................................................................... 279
Questions .......................................................................................... 279
281
Task 13.1: Launching emacs and Inserting Text ................................ 282
Task 13.2: How To Move Around in a File ....................................... 285
Task 13.3: How To Delete Characters and Words ............................ 289
Task 13.4: Search and Replace in emacs ............................................ 294
Task 13.5: Using the emacs Tutorial and Help System ...................... 297
Task 13.6: Working with Other Files ................................................ 299
Summary ................................................................................................ 303
Workshop .............................................................................................. 303
Key Terms ......................................................................................... 303
Questions .......................................................................................... 303
305
Task 14.1: What Shells Are Available? ............................................... 306
Task 14.2: Identifying Your Shell ...................................................... 309
Task 14.3: How To Choose a New Shell ........................................... 310
Task 14.4: Learning the Shell Environment ....................................... 313
Task 14.5: Exploring csh Configuration Files ................................... 317
Summary ................................................................................................ 321
Workshop .............................................................................................. 321
Key Terms ......................................................................................... 321
Questions .......................................................................................... 321
Contents
15
xi
323
Task 15.1: The C Shell and Korn Shell History Mechanisms ............ 324
Task 15.2: Using History to Cut Down on Typing ........................... 327
Task 15.3: Command Aliases ............................................................. 333
Task 15.4: Some Power Aliases .......................................................... 335
Task 15.5: Setting Custom Prompts .................................................. 338
Task 15.6: Creating Simple Shell Scripts ........................................... 340
Summary ................................................................................................ 344
Workshop .............................................................................................. 344
Key Terms ......................................................................................... 344
Questions .......................................................................................... 344
16
17
18
347
Task 16.1: Shell Variables .................................................................. 348
Task 16.2: Shell Arithmetic ............................................................... 350
Task 16.3: Comparison Functions ..................................................... 351
Task 16.4: Conditional Expressions ................................................... 355
Task 16.5: Looping expressions ......................................................... 357
Summary ................................................................................................ 359
Workshop .............................................................................................. 359
Key Terms ......................................................................................... 360
Questions .......................................................................................... 360
Job Control
361
Task 17.1: Job Control in the Shell: Stopping Jobs ............................ 362
Task 17.2: Foreground/Background and UNIX Programs ................. 365
Task 17.3: Finding Out What Tasks Are Running ............................ 368
Task 17.4: Terminating Processes with kill ..................................... 374
Summary ................................................................................................ 377
Workshop .............................................................................................. 377
Key Terms ......................................................................................... 377
Questions .......................................................................................... 378
379
Task 18.1: Find Local Printers with printers .................................. 380
Task 18.2: Printing Files with lpr or lp ............................................ 384
Task 18.3: Formatting Print Jobs with pr .......................................... 387
Task 18.4: Working with the Print Queue ........................................ 391
xii
Summary ................................................................................................ 394
Workshop .............................................................................................. 394
Key Terms ......................................................................................... 395
Questions .......................................................................................... 395
19
20
21
22
397
Task 19.1: The find Command and Its Weird Options .................... 398
Task 19.2: Using find with xargs .................................................... 403
Summary ................................................................................................ 405
Workshop .............................................................................................. 405
Questions .......................................................................................... 405
407
Task 20.1: Enabling Messages Using mesg ......................................... 408
Task 20.2: Writing to Other Users with write .................................. 409
Task 20.3: Reading Electronic Mail with mailx ................................ 411
Task 20.4: Sending Mail with mailx ................................................. 417
Task 20.5: The Smarter Electronic Mail Alternative, elm .................. 420
Summary ................................................................................................ 423
Workshop .............................................................................................. 423
Key Terms ......................................................................................... 424
Questions .......................................................................................... 424
425
Introduction to the Internet ................................................................... 426
Task 21.1: Starting Your Browser ...................................................... 427
Task 21.2: Finding Some Sites ........................................................... 432
Task 21.3: Customizing Your Browser ............................................... 437
Summary ................................................................................................ 440
Workshop .............................................................................................. 440
Key Terms ......................................................................................... 440
Questions .......................................................................................... 441
443
Task 22.1: Sending E-Mail to Internet Users ..................................... 444
Task 22.2: Talking with Remote Internet Users ................................ 446
Task 22.3: Searching Databases with WAIS ....................................... 449
Task 22.4: Having the Whole World with gopher ............................ 454
Contents
xiii
Task 22.5: Visiting Libraries Around the World ................................ 460
Task 22.6: All the News That’s Fit or Otherwise ............................... 466
Workshop .............................................................................................. 477
Key Terms ......................................................................................... 477
Questions .......................................................................................... 477
23
24
479
Task 23.1: Connecting to Remote Internet Sites ................................ 480
Task 23.2: Copying Files from Other Internet Sites ........................... 483
Task 23.3: Finding Archives with archie .......................................... 493
Task 23.4: A Few Interesting telnet Sites ........................................ 499
Workshop .............................................................................................. 507
Key Terms ......................................................................................... 507
Questions .......................................................................................... 507
509
Task 24.1: Your First Program ........................................................... 510
Task 24.2: Basic Data Types and Operators ...................................... 512
Task 24.3: Conditional Statements .................................................... 517
Task 24.4: Looping Statements .......................................................... 520
Task 24.5: Functions ......................................................................... 521
Task 24.6: Arrays ............................................................................... 523
Task 24.7: Pointers ............................................................................ 524
Task 24.8: Structures ......................................................................... 526
Summary ................................................................................................ 528
Where To Go Next ................................................................................ 528
Workshop .............................................................................................. 529
Key Terms ......................................................................................... 529
Questions .......................................................................................... 530
Glossary
531
Index
541
Contents
Teach Yourself
UNIX in 24 Hours
xiv
About the Authors
Dave Taylor
Dave Taylor is President and Chief Technical Officer of The Internet Mall, Inc., (http://
www.internetmall.com), the largest online shopping site in the world. He has been involved
with UNIX and the Internet since 1980, having created the popular Elm Mail System and
Embot mail autoresponder. A prolific author, he has been published over 1,000 times, and
his most recent books include the best-selling Creating Cool HTML 3.2 Web Pages and The
Internet Business Guide. Dave has a weekly intranet column in InfoWorld and a Web/CGI
programming column in LOGIN.
Previous positions include being a Research Scientist at HP Laboratories and Senior Reviews
Editor of SunWorld magazine. He also has contributed software to the official 4.4 release of
Berkeley UNIX (BSD), and his programs are found in all versions of Linux and other popular
UNIX variants.
Dave has a Bachelor’s degree in Computer Science (U.C.S.D., 1984) and a Master’s degree
in Education (Purdue, 1995), and he teaches evening courses in San Jose State University’s
Professional Development Program. His official home page on the Web is http://
www.intuitive.com/taylor, and his e-mail address for the last decade has been
[email protected].
James C. Armstrong, Jr.
James C. Armstrong, Jr., is the Director of Engineering at The Internet Mall, Inc., a San Jose,
California-based firm, dedicated to making Web-based commerce a turnkey operation.
James has nearly 15 years of professional experience with UNIX software products and has
worked for Bell Labs, Sun, and Tandem Computers in the past. He is also an 18-year veteran
of the Internet and its predecessors; his first contact was as a college student, exchanging
electronic mail with his father at AT&T.
James has a Bachelor’s degree in Computer Science from Duke University and has done some
graduate study at the University of St. Andrews in Scotland. James is an avid naturalist and
environmentalist and has traveled the world to photograph the beauty of nature.
xiv
Tell Us What You Think!
As a reader, you are the most important critic and commentator of our books. We value your
opinion and want to know what we’re doing right, what we could do better, what areas you’d
like to see us publish in, and any other words of wisdom you’re willing to pass our way. You
can help us make strong books that meet your needs and give you the computer guidance you
require.
Do you have access to CompuServe or the World Wide Web? Then check out our
CompuServe forum by typing GO SAMS at any prompt. If you prefer the World Wide Web,
check out our site at http://www.mcp.com.
If you have a technical question about this book, call the technical support
line at 317-581-4669.
JUST A MINUTE
As the team leader of the group that created this book, I welcome your comments. You can
fax, e-mail, or write me directly to let me know what you did or didn’t like about this book—
as well as what we can do to make our books stronger. Here’s the information:
Fax:
317-581-4669
E-mail:
[email protected]
Mail:
Dean Miller
Comments Department
Sams Publishing
201 W. 103rd Street
Indianapolis, IN 46290
xvi
Introduction
Welcome to Teach Yourself UNIX in 24 Hours ! This book has been designed so it is helpful
for both beginning users and those with previous UNIX experience. This text is helpful as
a guide, as well as a tutorial. The reader of this book is assumed to be intelligent, but no
familiarity with UNIX is expected.
Does Each Chapter Take an Hour?
You can learn the concepts in each of the 24 chapters in one hour. If you want to experiment
with what you learn in each chapter, you may take longer than an hour. However, all the
concepts presented here are straightforward. If you are familiar with Windows applications,
you will be able to progress more quickly through it.
How To Use This Book
This book is designed to teach you topics in one-hour sessions. All the books in the Sams
Teach Yourself series enable you to start working and become productive with the product
as quickly as possible. This book will do that for you!
Each hour, or session, starts with an overview of the topic to inform you what to expect in
each lesson. The overview helps you determine the nature of the lesson and whether the lesson
is relevant to your needs.
Main Section
Each lesson has a main section that discusses the lesson topic in a clear, concise manner by
breaking the topic down into logical component parts and explaining each component
clearly.
Interspersed in each lesson are special elements, called Just a Minutes, Time Savers, and
Cautions, that provide additional information.
JUST A MINUTE
Just a Minutes are designed to clarify the concept that is being discussed.
It elaborates on the subject, and if you are comfortable with your understanding of the subject, you can bypass them without danger.
TIME SAVER
CAUTION
Time Savers inform you of tricks or elements that are easily missed by most
computer users. You can skip them, but often Time Savers show you an
easier way to do a task.
A Caution deserves at least as much attention as a Time Saver because
Cautions point out a problematic element of the topic being discussed.
Ignoring the information contained in the Caution could have adverse
effects on the task at hand. These are the most important special elements
in this book.
Tasks
This book offers another special element called a Task. These step-by-step exercises are
designed to quickly walk you through the most important skills you can learn in UNIX. Each
Task has three parts—Description, Action, and Summary.
Workshops
The Workshop section at the end of each lesson provides Key Terms and Questions that
reinforce concepts you learned in the lesson and help you apply them in new situations. You
can skip this section, but it is advised that you go through the exercises to see how the concepts
can be applied to other common tasks. The Key Terms also are compiled in one alphabetized
list in the Glossary at the end of the book.
1
1
Hour
1
What Is This UNIX
Stuff?
Welcome to Teach Yourself UNIX in 24 Hours! This hour starts you toward
becoming a UNIX expert. Our goal for the first hour is to introduce you to some
UNIX history and to teach you where to go for help online.
Goals for This Hour
In the first hour, you learn
■
■
■
■
■
■
■
The history of UNIX
Why it’s called UNIX
What multiuser systems are all about
The difference between UNIX and other operating systems
About command-line interpreters and how users interact with UNIX
How to use man pages, UNIX’s online reference material
Other ways to find help in UNIX
Hour 1
2
What Is UNIX?
UNIX is a computer operating system, a control program that works with users to run
programs, manage resources, and communicate with other computer systems. Several people
can use a UNIX computer at the same time; hence UNIX is called a multiuser system. Any
of these users can also run multiple programs at the same time; hence UNIX is called
multitasking. Because UNIX is such a pastiche—a patchwork of development—it’s a lot
more than just an operating system. UNIX has more than 250 individual commands. These
range from simple commands—for copying a file, for example—to the quite complex: those
used in high-speed networking, file revision management, and software development.
Most notably, UNIX is a multichoice system. As an example, UNIX has three different
primary command-line-based user interfaces (in UNIX, the command-line user interface is
called a shell ): The three choices are the Bourne shell, C shell, and Korn shell. Often, soon
after you learn to accomplish a task with a particular command, you discover there’s a second
or third way to do that task. This is simultaneously the greatest strength of UNIX and a source
of frustration for both new and current users.
Why is having all this choice such a big deal? Think about why Microsoft MS-DOS and the
Apple Macintosh interfaces are considered so easy to use. Both are designed to give the user
less power. Both have dramatically fewer commands and precious little overlap in commands:
You can’t use copy to list your files in DOS, and you can’t drag a Mac file icon around to
duplicate it in its own directory. The advantage to these interfaces is that, in either system,
you can learn the one-and-only way to do a task and be confident that you’re as sophisticated
in doing that task as is the next person. It’s easy. It’s quick to learn. It’s exactly how the experts
do it, too.
UNIX, by contrast, is much more like a spoken language, with commands acting as verbs,
command options (which you learn about later in this lesson) acting as adjectives, and the
more complex commands acting akin to sentences. How you do a specific task can, therefore,
be completely different from how your UNIX-expert friend does the same task. Worse, some
specific commands in UNIX have many different versions, partly because of the variations
from different UNIX vendors. (You’ve heard of these variations and vendors, I’ll bet:
UNIXWare from Novell, Solaris from Sun, SCO from Santa Cruz, System V Release 4
(pronounce that “system five release four” or, to sound like an ace, “ess-vee-are-four”), and
BSD UNIX (pronounced “bee-ess-dee”) from University of California at Berkeley are the
primary players. Each is a little different from the other.) Another contributor to the sprawl
of modern UNIX is the energy of the UNIX programming community; plenty of UNIX users
decide to write a new version of a command in order to solve slightly different problems, thus
spawning many versions of a command.
1
JUST A MINUTE
I must admit that I, too, am guilty of rewriting a variety of UNIX commands, including those for an electronic mail system, a simple lineoriented editor, a text formatter, a programming language interpreter,
calendar manager, and even slightly different versions of the file-listing
command ls and the remove-files command rm. As a programmer, I found
that trying to duplicate the functionality of a particular command or utility
was a wonderful way to learn more about UNIX and programming.
Given the multichoice nature of UNIX, I promise to teach you the most popular UNIX
commands, and, if there are alternatives, I will teach you about those, too. The goal of this
book is for you to learn UNIX and to be able to work alongside long-time UNIX folk as a
peer, sharing your expertise with them and continuing to learn about the system and its
commands from them and other sources.
A Brief History of UNIX
To understand why the UNIX operating system has so many commands and why it’s not only
the premier multiuser, multitasking operating system, but also the most successful and the
most powerful multichoice system for computers, you’ll have to travel back in time. You’ll
need to learn where UNIX was designed, what were the goals of the original programmers,
and what has happened to UNIX in the subsequent decades.
Unlike DOS, Windows, OS/2, the Macintosh, VMS, MVS, and just about any other
operating system, UNIX was designed by a couple of programmers as a fun project, and it
evolved through the efforts of hundreds of programmers, each of whom was exploring his or
her own ideas of particular aspects of OS design and user interaction. In this regard, UNIX
is not like other operating systems, needless to say!
It all started back in the late 1960s in a dark and stormy laboratory deep in the recesses of the
American Telephone and Telegraph (AT&T) corporate facility in New Jersey. Working with
the Massachusetts Institute of Technology, AT&T Bell Labs was codeveloping a massive,
monolithic operating system called Multics. On the Bell Labs team were Ken Thompson,
Dennis Ritchie, Brian Kernighan, and other people in the Computer Science Research
Group who would prove to be key contributors to the new UNIX operating system.
When 1969 rolled around, Bell Labs was becoming increasingly disillusioned with Multics,
an overly slow and expensive system that ran on General Electric mainframe computers that
themselves were expensive to run and rapidly becoming obsolete. The problem was that
Thompson and the group really liked the capabilities Multics offered, particularly the
individual-user environment and multiple-user aspects.
3
1
Hour 1
4
In that same year, Thompson wrote a computer game called Space Travel, first on Multics,
then on the GECOS (GE computer operating system). The game was a simulation of the
movement of the major bodies of the Solar System, with the player guiding a ship, observing
the scenery, and attempting to land on the various planets and moons. The game wasn’t much
fun on the GE computer, however, because performance was jerky and irregular, and, more
importantly, it cost almost $100 in computing time for each game.
In his quest to improve the game, Thompson found a little-used Digital Equipment
Corporation PDP-7, and with some help from Ritchie, he rewrote the game for the
PDP-7. Development was done on the GE mainframe and hand-carried to the PDP-7 on
paper tape.
Once he’d explored some of the capabilities of the PDP-7, Thompson couldn’t resist
building on the game, starting with an implementation of an earlier file system he’d designed,
then adding processes, simple file utilities (cp, mv), and a command interpreter that he called
a “shell.” It wasn’t until the following year that the newly created system acquired its name,
UNIX, which Brian Kernighan suggested as a pun on Multics.
The Thompson file system was built around the low-level concept of i-nodes—linked blocks
of information that together comprise the contents of a file or program—kept in a big list
called the i-list, subdirectories, and special types of files that described devices and acted as
the actual device driver for user interaction. What was missing in this earliest form of UNIX
was pathnames. No slash (/) was present, and subdirectories were referenced through a
confusing combination of file links that proved too complex, causing users to stop using
subdirectories. Another limitation in this early version was that directories couldn’t be added
while the system was running and had to be added to the preload configuration.
In 1970, Thompson’s group requested and received a Digital PDP-11 system for the purpose
of creating a system for editing and formatting text. It was such an early unit that the first disk
did not arrive at Bell Labs until four months after the CPU showed up. The first important
program on UNIX was the text-formatting program roff, which—keep with me now—was
inspired by McIlroy’s BCPL program on Multics, which in turn had been inspired by an
earlier program called runoff on the CTSS operating system.
The initial customer was the Patent Department inside the Labs, a group that needed a system
for preparing patent applications. There, UNIX was a dramatic success, and it didn’t take
long for others inside Bell Labs to begin clamoring for their own UNIX computer systems.
The C Programming Language
That’s where UNIX came from. What about C, the programming language that is integral
to the system?
1
In 1969, the original UNIX had a very-low-level assembly language compiler available for
writing programs; all the PDP-7 work was done in this primitive language. Just before the
PDP-11 arrived, McIlroy ported a language called TMG to the PDP-7, which Thompson
then tried to use to write a FORTRAN compiler. That didn’t work, and instead he produced
a language called B. Two years later, in 1971, Ritchie created the first version of a new
programming language based on B, a language he called C. By 1973, the entire UNIX system
had been rewritten in C for portability and speed.
UNIX Becomes Popular
In the 1970s, AT&T hadn’t yet been split up into the many regional operating companies
known today, and the company was prohibited from selling the new UNIX system. Hoping
for the best, Bell Labs distributed UNIX to colleges and universities for a nominal charge.
These institutions also were happily buying the inexpensive and powerful PDP-11 computer
systems—a perfect match. Before long, UNIX was the research and software-development
operating system of choice.
The UNIX of today is not, however, the product of a couple of inspired programmers at Bell
Labs. Many other organizations and institutions contributed significant additions to the
system as it evolved from its early beginnings and grew into the monster it is today. Most
important were the C shell, TCP/IP networking, vi editor, Berkeley Fast File System, and
sendmail electronic-mail-routing software from the Computer Science Research Group of
the University of California at Berkeley. Also important were the early versions of UUCP and
Usenet from the University of Maryland, Delaware, and from Duke University. After
dropping Multics development completely, MIT didn’t come into the UNIX picture until
the early 1980s, when it developed the X Window System as part of its successful Athena
project. Ten years and four releases later, X is the predominant windowing system standard
on all UNIX systems, and it is the basis of Motif, OpenWindows, and Open Desktop.
Gradually, big corporations have become directly involved with the evolutionary process,
notably Hewlett-Packard, Sun Microsystems, and Digital Equipment Corporation. Little
companies have started to get into the action too, with UNIX available from Apple for the
Macintosh and from IBM for PCs, RISC-based workstations, and new PowerPC computers.
Today, UNIX runs on all sizes of computers, from humble PC laptops, to powerful desktopvisualization workstations, and even to supercomputers that require special cooling fluids to
prevent them from burning up while working. It’s a long way from Space Travel, a game that,
ironically, isn’t part of UNIX anymore.
What’s All This About Multiuser Systems?
Among the many multi words you learned earlier was one that directly concerns how you
interact with the computer, multiuser. The goal of a multiuser system is for all users to feel
5
1
Hour 1
6
as though they’ve each been given their own personal computer, their own individual UNIX
system, although they actually are working within a large system. To accomplish this, each
user is given an account—usually based on the person’s last name, initials, or another unique
naming scheme—and a home directory, the default place where his or her files are saved. This
leads to a bit of a puzzle: When you’re working on the system, how does the system know
that you’re you? What’s to stop someone else from masquerading as you, going into your files,
prying into private letters, altering memos, or worse?
On a Macintosh or PC, anyone can walk up to your computer when you’re not around, flip
the power switch, and pry, and you can’t do much about it. You can add some security
software, but security isn’t a fundamental part of the system, which results in an awkward fit
between system and software. For a computer sitting on your desk in your office, though,
that’s okay; the system is not a shared multiuser system, so verifying who you are when you
turn on the computer isn’t critical.
But UNIX is a system designed for multiple users, so it is very important that the system can
confirm your identity in a manner that precludes others from masquerading as you. As a
result, all accounts have passwords associated with them—like a PIN for a bank card, keep
it a secret!—and, when you use your password in combination with your account, the
computer can be pretty sure that you are who you’re claiming to be. For obvious reasons,
when you’re done using the computer, you always should remember to end your session, or,
in effect, to turn off your virtual personal computer when you’re done.
In the next hour, you learn your first UNIX commands. At the top of the list are commands
to log in to the system, enter your password, and change your password to be memorable and
highly secure.
Cracking Open the Shell
Another unusual feature of UNIX systems, especially for those of you who come from either
the Macintosh or the Windows environments, is that UNIX is designed to be a command–
line-based system rather than a more graphically based (picture-oriented) system. That’s a
mixed blessing. It makes UNIX harder to learn, but the system is considerably more powerful
than fiddling with a mouse to drag little pictures about on the screen.
There are graphical interfaces to UNIX, built within the X Window System environment.
Notable ones are Motif, Open Windows, and Open Desktop. Even with the best of these,
however, the command-line heart of UNIX still shines through, and in my experience, it’s
impossible really to use all the power that UNIX offers without turning to a shell.
If you’re used to writing letters to your friends and family or even mere shopping lists, you
won’t have any problem with a command-line interface: It’s a command program that you
tell what to do. When you type specific instructions and press the Return key, the computer
leaps into action and immediately performs whatever command you’ve specified.
1
JUST A MINUTE
Throughout this book, I refer to pressing the Return key, but your keyboard
may have this key labeled as “Enter” or marked with a left-pointing,
specially shaped arrow. These all mean the same thing.
In Windows, you might move a file from one folder to another by opening the folder, opening
the destination folder, fiddling around for a while to be sure that you can see both of them
on the screen at the same time, and then clicking and dragging the specific file from one place
to the other. In UNIX it’s much easier. Typing in the following simple command does the trick:
cp folder1/file folder2
It automatically ensures the file has the same name in the destination directory, too.
This might not seem much of a boon, but imagine the situation where you want to move all
files with names that start with the word project or end with the suffix .c (C program files).
This could be quite tricky and could take a lot of patience with a graphical interface. UNIX,
however, makes it easy:
cp project* *.c folder2
Soon you not only will understand this command, but you also will be able to compose your
own examples!
Getting Help
Throughout this book, the focus is on the most important and valuable flags and options for
the commands covered. That’s all well and good, but how do you find out about the other
alternatives that might actually work better for your use? That’s where the UNIX “man” pages
come in. You will learn how to browse them to find the information desired.
Task 1.1: Man Pages, UNIX Online Reference
It’s not news to you that UNIX is a very complex operating system, with hundreds
of commands that can be combined to execute thousands of possible actions. Most
commands have a considerable number of options, and all seem to have some subtlety or
other that it’s important to know. But how do you figure all this out? You need to look up
commands in the UNIX online documentation set. Containing purely reference materials,
the UNIX man pages (man is short for manual ) cover every command available.
To search for a man page, enter man followed by the name of the command to find. Many
sites also have a table of contents for the man pages (it’s called a whatis database, for obscure
historical reasons.) You can use the all-important -k flag for keyword searches, to find the
name of a command if you know what it should do but you just can’t remember what it’s
called.
7
1
Hour 1
8
JUST A MINUTE
A command performs a basic task, which can be modified by adding
flags to the end of the command when you enter it on the command line.
These flags are described in the man pages. For example, to use the –k
flag for man, enter:
% man –k
JUST A MINUTE
The command apropos is available on most UNIX systems and is often just
an alias to man -k. If it’s not on your system, you can create it by adding
the following line to your .cshrc file:
alias apropos ‘man -k \!’
The UNIX man pages are organized into nine sections, as shown in Table 1.1. This table is
organized for System V, but it generally holds true for Berkeley systems, too, with these few
changes: BSD has I/O and special files in Section 4, administrative files in Section 5, and
miscellaneous files in Section 7. Some BSD systems also split user commands into further
categories: Section 1C for intersystem communications and Section 1G for commands used
primarily for graphics and computer-aided design.
Table 1.1. System V UNIX man page organization.
1
Section
Category
1
1M
2
3
4
5
6
7
8
User commands
System maintenance commands
System calls
Library routines
Administrative files
Miscellaneous
Games
I/O and special files
Administrative commands
9
1
1. The mkdir man page is succinct and exemplary:
% man mkdir
MKDIR(1)
DYNIX Programmer’s Manual
MKDIR(1)
NAME
mkdir - make a directory
SYNOPSIS
mkdir dirname ...
DESCRIPTION
Mkdir creates specified directories in mode 777. Standard
entries, `.’, for the directory itself, and `..’ for its
parent, are made automatically.
Mkdir requires write permission in the parent directory.
SEE ALSO
rmdir(1)
Revision 1.4.2.2 88/08/13
%
JUST A MINUTE
1
Notice in the example, that in the first line, the command itself is in
boldface type, but everything else is not bold. Throughout this book,
whenever an example contains both user input and UNIX output, the user
input will be bold so that you can spot easily what you are supposed to
enter.
The very first line of the output tells me that it’s found the mkdir command in
Section 1 (user commands) of the man pages, with the middle phrase, DYNIX
Programmer’s Manual , indicating that I’m running on a version of UNIX called
DYNIX. The NAME section always details the name of the command and a one-line
summary of what it does. SYNOPSIS explains how to use the command, including all
possible command flags and options.
DESCRIPTION is where all the meaningful information is, and it can run on for
dozens of pages, explaining how complex commands like csh or vi work. SEE ALSO
suggests other commands that are related in some way. The Revision line at the
bottom is different on each version of man, and it indicates the last time, presumably, that this document was revised.
Hour 1
10
2. The same man page from a Sun workstation is quite different:
% man mkdir
MKDIR(1)
USER COMMANDS
MKDIR(1)
NAME
mkdir - make a directory
SYNOPSIS
mkdir [ -p ] dirname...
DESCRIPTION
mkdir creates directories. Standard entries, `.’, for the
directory itself, and `..’ for its parent, are made automatically.
The -p flag allows missing parent directories to be
as needed.
created
With the exception of the set-gid bit, the current umask(2V)
setting
determines the mode in which directories are
created. The new directory inherits the set-gid bit of the
parent directory.
Modes may be modified after creation by
using chmod(1V).
mkdir requires write permission in the parent directory.
SEE ALSO
chmod(1V), rm(1), mkdir(2V), umask(2V)
Sun Release 4.1
%
Last change: 22 August 1989
1
Notice that there’s a new flag in this version of mkdir, the -p flag. More importantly, note that the flag is shown in square brackets within the SYNOPSIS section.
By convention, square brackets in this section mean that the flag is optional. You
can see that the engineers at Sun have a very different idea about what other
commands might be worth viewing!
3. One thing I always forget on Sun systems is the command that lets me format a
floppy disk. That’s exactly where the apropos command comes in handy:
% apropos floppy
fd (4S)
%
- disk driver for Floppy Disk Controllers
That’s not quite what I want, unfortunately. Because it’s in Section 4 (note that the
word in parentheses is 4S, not 1), this document will describe the disk driver rather
than any command to work with floppy disks.
I can look up the disk command instead:
% man -k disk
acctdisk, acctdusg, accton, acctwtmp (8)
- overview of accounting and
➥miscellaneous accounting commands
add_client (8)
- create a diskless network bootable NFS client on
1
➥a server
chargefee, ckpacct, dodisk, lastlogin, monacct, nulladm, prctmp, prdaily,
➥prtacct, runacct, shutacct, startup, turnacct (8) - shell procedures for
➥accounting
client (8)
- add or remove diskless Sun386i systems
df (1V)
- report free disk space on file systems
diskusg (8)
- generate disk accounting data by user
dkctl (8)
- control special disk operations
dkinfo (8)
- report information about a disk’s geometry and
➥partitioning
dkio (4S)
- generic disk control operations
du (1L)
- summarize disk usage
du (1V)
- display the number of disk blocks used per
➥directory or file
fastboot, fasthalt (8) - reboot/halt the system while disabling disk
➥checking
fd (4S)
- disk driver for Floppy Disk Controllers
fdformat (1)
- format diskettes for use with SunOS
format (8S)
- disk partitioning and maintenance utility
fsync (2)
- synchronize a file’s in-core state with that
➥on disk
fusage (8)
- RFS disk access profiler
id (4S)
- disk driver for IPI disk controllers
installboot (8S)
- install bootblocks in a disk partition
pnpboot, pnp.s386 (8C) - pnp diskless boot service
quota (1)
- display a user’s disk quota and usage
quotactl (2)
- manipulate disk quotas
root (4S)
- pseudo-driver for Sun386i root disk
sd (4S)
- driver for SCSI disk devices
sync (1)
- update the super block; force changed blocks
➥to the disk
xd (4S)
- Disk driver for Xylogics 7053 SMD Disk
➥Controller
xy (4S)
- Disk driver for Xylogics 450 and 451 SMD Disk
➥Controllers
%
JUST A MINUTE
Notice the ➥ character at the beginning of some of the lines in this
example. This character does not appear on your screen. It’s a typographical convention used in the book because the number of characters
that can be displayed by UNIX on a line of your screen is greater than the
number of characters that can appear (legibly) on a line in this book. The
➥ indicates that the text following it is actually part of the preceding line
on your screen.
This yields quite a few choices! To trim the list down to just those that are in
Section 1 (the user commands section), I use grep:
% man -k disk | grep ‘(1’
df (1V)
- report free disk space on file systems
du (1L)
- summarize disk usage
11
1
Hour 1
12
du (1V)
➥directory or file
fdformat (1)
quota (1)
sync (1)
➥to the disk
%
- display the number of disk blocks used per
- format diskettes for use with SunOS
- display a user’s disk quota and usage
- update the super block; force changed blocks
That’s better! The command I was looking for is fdformat.
4. To learn a single snippet of information about a UNIX command, you can check
to see if your system has the whatis utility. You can even ask it to describe itself (a
bit of a philosophical conundrum):
% whatis whatis
whatis (1)
%
- display a one-line summary about a keyword
In fact, this is the line from the NAME section of the relevant man page. The whatis
command is different from the apropos command because it considers only
command names rather than all words in the command description line:
% whatis cd
cd (1)
%
- change working directory
Now see what apropos does:
% apropos cd
bcd, ppt (6)
- convert to antique media
cd (1)
cdplayer (6)
- CD-ROM audio demo program
cdromio (4S)
- CDROM control operations
draw, bdraw, cdraw (6) - interactive graphics drawing
fcdcmd, fcd (1)
- change client’s current working directory in
➥the FSP database
getacinfo, getacdir, getacflg, getacmin, setac, endac (3)
- get audit
➥ control file information
ipallocd (8C)
- Ethernet-to-IP address allocator
mp, madd, msub, mult, mdiv, mcmp, min, mout, pow, gcd, rpow, itom, xtom,
➥mtox, mfree (3X)
- multiple precision integer arithmetic
rexecd, in.rexecd (8C) - remote execution server
sccs-cdc, cdc (1)
- change the delta commentary of an SCCS delta
sr (4S)
- driver for CDROM SCSI controller
termios, tcgetattr, tcsetattr, tcsendbreak, tcdrain, tcflush, tcflow,
➥cfgetospeed, cfgetispeed, cfsetispeed, cfsetospeed (3V) - get and
➥set terminal attributes, line control, get and set baud rate, get
➥and set terminal foreground process group ID
tin, rtin, cdtin, tind (1) - A threaded Netnews reader
uid_allocd, gid_allocd (8C) - UID and GID allocator daemons
%
5. One problem with man is that it really isn’t too sophisticated. As you can see in the
example in step 4, apropos (which, recall, is man -k) lists a line more than once if
more than one man page match the specified pattern. You can create your own
apropos alias to improve the command:
1
13
% alias apropos _man -k \!* | uniq_
% apropos cd
bcd, ppt (6)
- convert to antique media
cd (1)
cdplayer (6)
- CD-ROM audio demo program
cdromio (4S)
- CDROM control operations
draw, bdraw, cdraw (6) - interactive graphics drawing
fcdcmd, fcd (1)
- change client’s current working directory
➥in the FSP database
getacinfo, getacdir, getacflg, getacmin, setac, endac (3)
- get audit
➥control file information
ipallocd (8C)
- Ethernet-to-IP address allocator
mp, madd, msub, mult, mdiv, mcmp, min, mout, pow, gcd, rpow, itom, xtom,
➥mtox, mfree (3X)
- multiple precision integer arithmetic
rexecd, in.rexecd (8C) - remote execution server
sccs-cdc, cdc (1)
sr (4S)
- driver for CDROM SCSI controller
termios, tcgetattr, tcsetattr, tcsendbreak, tcdrain, tcflush, tcflow,
➥cfgetospeed, cfgetispeed, cfsetispeed, cfsetospeed (3V) - get and set
➥terminal attributes, line control, get and set baud rate, get
➥and set terminal foreground process group ID
uid_allocd, gid_allocd (8C) - UID and GID allocator daemons
%
That’s better, but I’d like to have the command tell me about only user commands
because I don’t care much about file formats, games, or miscellaneous commands
when I’m looking for a command. I’ll try this:
% alias apropos _man -k \!* | uniq | grep 1_
% apropos cd
cd (1)
fcdcmd, fcd (1)
- change client’s current working directory
➥in the FSP database
sccs-cdc, cdc (1)
%
That’s much better.
6. I’d like to look up one more command—sort—before I’m done here.
% man sort
SORT(1)
SORT(1)
NAME
sort - sort or merge files
SYNOPSIS
sort [ -mubdfinrtx ] [ +pos1 [ -pos2 ] ] ... [ -o name ]
[ -T directory ] [ name ] ...
DESCRIPTION
Sort sorts lines of all the named files together and writes
the result on the standard output. The name `-’ means the
standard input. If no input files are named, the standard
input is sorted.
1
Hour 1
14
The default sort key is an entire line. Default ordering is
lexicographic by bytes in machine collating sequence. The
ordering is affected globally by the following options, one
or more of which may appear.
b
Ignore leading blanks (spaces and tabs) in field com--More-- _
On almost every system, the man command feeds output through the more program so that
information won’t scroll by faster than you can read it. You also can save the output of a man
command to a file if you’d like to study the information in detail. To save this particular
manual entry to the file sort.manpage, you could use man sort > sort.manpage.
Notice in the sort man page that there are many options to the sort command (certainly
more than discussed in this book). As you learn UNIX, if you find areas about which you’d
like more information, or if you need a capability that doesn’t seem to be available, check the
man page. There just might be a flag for what you seek.
JUST A MINUTE
You can obtain lots of valuable information by reading the introduction to
each section of the man pages. Use man 1 intro to read the introduction
to Section 1, for example.
If your version of man doesn’t stop at the bottom of each page, you can
remedy the situation using alias man ‘man \!* | more’.
UNIX was one of the very first operating systems to include online documentation.
The man pages are an invaluable reference. Most of them are poorly written,
unfortunately, and precious few include examples of actual usage. However, as a quick
reminder of flags and options, or as an easy way to find out the capabilities of a command,
man is great. I encourage you to explore the man pages and perhaps even read the man page
on the man command itself.
Task 1.2: Other Ways to Find Help in UNIX
The man pages are really the best way to learn about what’s going on with UNIX
commands, but some alternatives also can prove helpful. Some systems have a help
command. Many UNIX utilities make information available with the -h or -? flag, too.
Finally, one trick you can try is to feed a set of gibberish flags to a command, which sometimes
generates an error and a helpful message reminding you what possible options the command
accepts.
1
15
1. At the University Tech Computing Center, the support team has installed a help
command:
% help
Look in a printed manual, if you can, for general help. You should have
someone show you some things and then read one of the tutorial papers
(e.g., UNIX for Beginners or An Introduction to the C Shell) to get
started. Printed manuals covering all aspects of Unix are on sale at the
bookstore.
Most of the material in the printed manuals is also available online
via “man” and similar commands; for instance:
apropos keyword - lists commands relevant to keyword
whatis filename - lists commands involving filename
man command - prints out the manual entry for a command
help command - prints out the pocket guide entry for a command
➥are helpful; other basic commands are:
cat - display a file on the screen
date - print the date and time
du - summarize disk space usage
edit - text editor (beginner)
ex - text editor (intermediate)
finger - user information lookup program
learn - interactive self-paced tutorial on Unix
--More(40%)-- _
Your system might have something similar.
2. Some commands offer helpful output if you specify the -h flag:
% ls -h
usage: ls [ -acdfgilqrstu1ACLFR ] name ...
%
Then again, others don’t:
% ls -h
Global.Software
Interactive.Unix
%
Mail/
News/
Src/
bin/
history.usenet.Z
testme
A few commands offer lots of output when you use the -h flag:
% elm -h
Possible Starting Arguments for ELM program:
arg
Meaning
-a
Arrow - use the arrow pointer regardless
-c
Checkalias - check the given aliases only
-dn
Debug - set debug level to ‘n’
-fx
Folder - read folder ‘x’ rather than incoming mailbox
-h
Help - give this list of options
1
Hour 1
16
-k
-K
-m
-sx
-V
-v
-w
-z
%
Keypad - enable HP 2622 terminal keyboard
Keypad&softkeys - enable use of softkeys + “-k”
Menu - Turn off menu, using more of the screen
Subject ‘x’ - for batchmailing
Enable sendmail voyeur mode.
Print out ELM version information.
Supress warning messages...
Zero - don’t enter ELM if no mail is pending
Unfortunately, there isn’t a command flag common to all UNIX utilities that lists
the possible command flags.
3. Sometimes you can obtain help from a program by incurring its wrath. You can
specify a set of flags that are impossible, unavailable, or just plain puzzling. I always
use -xyz because they’re uncommon flags:
% man -xyz
man: unknown option ‘-x’, use ‘-h’ for help
Okay, I’ll try it:
% man -h
man: usage [-S | -t | -w] [-ac] [-m path] [-M path] [section] pages
man: usage -k [-ac] [-m path] [-M path] [section] keywords
man: usage -f [-ac] [-m path] [-M path] [section] names
man: usage -h
man: usage -V
a
display all manpages for names
c
cat (rather than page) manual pages
f
find whatis entries for pages by these names
names
names to search for in whatis
h
print this help message
k
find whatis entries by keywords
keywords keywords to search for in whatis
m path
add to the standard man path directories
M path
override standard man path directories
S
display only SYNOPSIS section of pages
t
find the source (rather than the formatted page)
V
show version information
w
only output which pages we would display
section section for the manual to search
pages
pages to locate
%
For every command that does something marginally helpful, there are a half-dozen
commands that give useless, and amusingly different, output for these flags:
% bc -xyz
unrecognizable argument
% cal -xyz
Bad argument
% file -xyz
-xyz:
No such file or directory
% grep -xyz
grep: unknown flag
%
1
You can’t rely on programs to be helpful about themselves, but you can rely on the
man page being available for just about everything on the system.
As much as I’d like to tell you that there is a wide variety of useful and interesting
information available within UNIX on the commands therein, in reality, UNIX has
man pages but precious little else. Furthermore, some commands installed locally might not
even have man page entries, which leaves you to puzzle out how they work. If you encounter
commands that are undocumented, I recommend that you ask your system administrator or
vendor what’s going on and why there’s no further information on the program.
Some vendors are addressing this problem in innovative, if somewhat limited, ways. Sun
Microsystems, for example, offers its complete documentation set, including all tutorials,
user guides, and man pages, on a single CD-ROM. AnswerBook, as it’s called, is helpful but
has some limitations, not the least of which is that you must have a CD-ROM drive and keep
the disk in the drive at all times.
Summary
In this first hour, the goal was for you to learn a bit about what UNIX is, where it came from,
and how it differs from other operating systems that you might have used in the past. You
also learned about the need for security on a multiuser system and how a password helps
maintain that security, so that your files are never read, altered, or removed by anyone but
yourself.
You also learned what a command shell, or command-line interpreter, is all about, how it
differs from graphically oriented interface systems like the Macintosh and Windows, and
how it’s not only easy to use, but considerably more powerful than dragging-and-dropping
little pictures.
Finally, you learned about getting help on UNIX. Although there aren’t many options, you
do have the manual pages available to you, as well as the command-line arguments and
apropos.
Workshop
The Workshop summarizes the key terms you learned and poses some questions about the
topics presented in this chapter. It also provides you with a preview of what you will learn
in the next hour.
Key Terms
account This is the official one-word name by which the UNIX system knows you.
Mine is taylor.
17
1
Hour 1
18
arguments Not any type of domestic dispute, arguments are the set of options and
filenames specified to UNIX commands. When you use a command such as vi test.c, all
words other than the command name itself (vi) are arguments, or parameters to the program.
i-list See i-node.
i-node The UNIX file system is like a huge notebook full of sheets of information. Each
file is like an index tab, indicating where the file starts in the notebook and how many sheets
are used. The tabs are called i-nodes, and the list of tabs (the index to the notebook) is the
i-list.
command Each program in UNIX is also known as a command: the two words are
interchangeable.
man page Each standard UNIX command comes with some basic online documentation
that describes its function. This online documentation for a command is called a man page.
Usually, the man page lists the command-line flags and some error conditions.
multitasking A multitasking computer is one that actually can run more than one
program, or task, at a time. By contrast, most personal computers lock you into a single
program that you must exit before you launch another.
multiuser Computers intended to have more than a single person working on them
simultaneously are designed to support multiple users, hence the term multiuser. By contrast,
personal computers are almost always single-user because someone else can’t be running a
program or editing a file while you are using the computer for your own work.
pathname UNIX is split into a wide variety of different directories and subdirectories,
often across multiple hard disks and even multiple computers. So that the system needn’t
search laboriously through the entire mess each time you request a program, the set of
directories you reference are stored as your search path, and the location of any specific
command is known as its pathname.
shell To interact with UNIX, you type in commands to the command-line interpreter,
which is known in UNIX as the shell, or command shell. It’s the underlying environment in
which you work with the UNIX system.
Questions
Each hour concludes with a set of questions for you to contemplate. Here’s a warning up
front: Not all of the questions have a definitive answer. After all, you are learning about a
multichoice operating system!
1
1. Name the three multi concepts that are at the heart of UNIX’s power.
2. Is UNIX more like a grid of streets, letting you pick your route from point A to
point B, or a directed highway with only one option? How does this compare with
other systems you’ve used?
3. Systems that support multiple users always ask you to say who you are when you
begin using the system. What’s the most important thing to remember when you’re
done using the system?
4. If you’re used to graphical interfaces, try to think of a few tasks that you feel are
more easily accomplished by moving icons than by typing commands. Write those
tasks on a separate paper, and in a few days, pull that paper out and see if you still
feel that way.
5. Think of a few instances in which you needed to give a person written instructions.
Was that easier than giving spoken instructions or drawing a picture? Was it
harder?
Preview of the Next Hour
In the next hour, you learn how to log in to the system at the login prompt (login:), how
to log out of the system, how to use the passwd command to change your password, how to
use the id command to find out who the computer thinks you are, and lots more!
19
1
Hour
2
Getting onto the
System and Using the
Command Line
This is the second hour of UNIX lessons, so it’s time you logged in to the system
and tried some commands. This hour focuses on teaching you the basics of
interacting with your UNIX machine.
Goals for This Hour
In this hour, you learn how to
■
■
■
■
■
Log in and log out of the system
Change passwords with passwd
Choose a memorable and secure password
Find out who the computer thinks you are
Find out who else is on the system
21
2
Hour 2
22
■
■
■
■
Find out what everyone is doing on the system
Check the current date and time
Look at a month and year calendar
Perform some simple calculations with UNIX
This hour introduces a lot of commands, so it’s very important that you have a UNIX system
available on which you can work through all examples. Most examples have been taken from
a Sun workstation running Solaris, a variant of UNIX System V Release 4, and have been
double-checked on a BSD-based system. Any variance between the two is noted, and if you
have a UNIX system available, odds are good that it’s based on either AT&T System V or
Berkeley UNIX.
Task 2.1: Logging In and Out of the System
Because UNIX is a multiuser system, you need to start by finding a terminal,
computer, or other way to access the system. I use a Macintosh and a modem to dial
up various systems by telephone. You might have a similar approach, or you might have a
terminal directly connected to the UNIX computer on your desk or in your office, or you
might have the UNIX system itself on your desk. Regardless of how you connect to your
UNIX system, the first thing you’ll see on the screen is this:
4.3BSD DYNIX (mentor.utech.edu) 5:38pm on Fri, 7 Feb 1997
login:
The first line indicates what variant of UNIX the system is running (DYNIX is UNIX on
Sequent computers), the actual name of the computer system, and the current date and time.
The second line is asking for your login, your account name.
1. Connect your terminal or PC to the UNIX system until the point where you see a
login prompt (login:) on your screen similar to that in the preceding example. Use
the phone and modem to dial up the computer if you need to.
It would be nice if computers could keep track of us users by simply using our full
names so that I could enter Dave Taylor at the login prompt. Alas, like the Internal
Revenue Service, Department of Motor Vehicles, and many other agencies,
UNIX—rather than using names—assigns each user a unique identifier. This
identifier is called an account name, has eight characters or fewer, and is usually
based on the first or last name, although it can be any combination of letters and
numbers. I have two account names, or logins, on the systems I use: taylor and, on
another machine where someone already had that account name, dataylor.
2
2. You should know your account name on the UNIX system. Perhaps your account
name is on a paper with your initial password, both assigned by the UNIX system
administrator. If you do not have this information, you need to track it down
before you can go further. Some accounts might not have an initial password; that
means that you won’t have to enter one the first time you log in to the system. In a
few minutes, you will learn how you can give yourself the password of your choice
by using a UNIX command called passwd.
3. At the login prompt, enter your account name. Be particularly careful to use all
lowercase letters unless specified otherwise by your administrator.
login: taylor
Password:
Once you’ve entered your account name, the system moves the cursor to the next
line and prompts you for your password. When you enter your password, the
system won’t echo it (that is, won’t display it) on the screen. That’s okay. Lack of
an echo doesn’t mean anything is broken; instead, this is a security measure to
ensure that even if people are looking over your shoulder, they can’t learn your
secret password by watching your screen.
4. If you enter either your login or your password incorrectly, the system complains
with an error message:
login: taylor
Password:
Login incorrect
login:
CAUTION
Most systems give you three or four attempts to get both your login and
password correct, so try again. Don’t forget to enter your account name at
the login prompt each time.
5. Once you’ve successfully entered your account name and password, you are shown
some information about the system, some news for users, and an indication of
whether you have electronic mail. The specifics will vary, but here’s an example of
what I see when I log in to my account:
login: taylor
Password:
Last login: Fri Feb 7 17:00:23 on ttyAe
You have mail.
%
23
2
Hour 2
24
JUST A MINUTE
The percent sign is UNIX’s way of telling you that it’s ready for you to
enter some commands. The percent sign is the equivalent of an enlisted
soldier saluting and saying, “Ready for duty!” or an employee saying,
“What shall I do now, boss?”
Your system might be configured so that you have some slightly different prompt
here. The possibilities include a $ for the Korn or Bourne shells, your current
location in the file system, the current time, the command-index number (which
you’ll learn about when you learn how to teach the UNIX command-line interpreter to adapt to your work style, rather than vice versa), and the name of the
computer system itself. Here are some examples:
[/users/taylor] :
(mentor) 33 :
taylor@mentor %
Your prompt might not look exactly like any of these, but it has one unique
characteristic: it is at the beginning of the line that your cursor sits on, and it
reappears each time you’ve completed working with any UNIX program.
6. At this point, you’re ready to enter your first UNIX command—exit—to sign off
from the computer system. Try it. On my system, entering exit shuts down all my
programs and hangs up the telephone connection. On other systems, it returns the
login prompt. Many UNIX systems offer a pithy quote as you leave, too.
% exit
He who hesitates is lost.
4.3BSD DYNIX (mentor.utech.edu) 5:38pm on Fri, 7 Feb 1993
login:
CAUTION
UNIX is case-sensitive, so the exit command is not the same as EXIT. If
you enter a command all in uppercase, the system won’t find it and
instead will respond with the complaint command not found.
7. If you have a direct connection to the computer, odds are very good that logging
out causes the system to prompt for another account name, enabling the next
person to use the system. If you dialed up the system with a modem, you probably
will see something more like the following example. After being disconnected,
you’ll be able to shut down your computer.
% exit
Did you lose your keys again?
DISCONNECTED
2
25
At this point, you’ve overcome the toughest part of UNIX. You have an account,
know the password, logged in to the system, and entered a simple command telling
the computer what you want to do, and the computer has done it!
Task 2.2: Changing Passwords with passwd
Having logged in to a UNIX system, you can clearly see that there are many
differences between UNIX and a PA or Macintosh personal computer. Certainly the
style of interaction is different. With UNIX, the keyboard becomes the exclusive method of
instructing the computer what to do, and the mouse sits idle, waiting for something to
happen.
One of the greatest differences is that UNIX is a multiuser system, as you learned in the
previous hour. As you learn more about UNIX, you’ll find that this characteristic has an
impact on a variety of tasks and commands. The next UNIX command you learn is one that
exists because of the multiuser nature of UNIX: passwd.
With the passwd command, you can change the password associated with your individual
account name. As with the personal identification number (PIN) for your automated-teller
machine, the value of your password is directly related to how secret it remains.
CAUTION
UNIX is careful about the whole process of changing passwords. It
requires you to enter your current password to prove you’re really you.
Imagine that you are at a computer center and have to leave the room to
make a quick phone call. Without much effort, a prankster could lean over
and quickly change your password to something you wouldn’t know.
That’s why you should log out if you’re not going to be near your system,
and that’s also why passwords are never echoed in UNIX.
1. Consider what happens when I use the passwd command to change the password
associated with my account:
% passwd
Changing password for taylor.
Old password:
New passwd:
Retype new passwd:
%
2. Notice that I never received any visual confirmation that the password I actually
entered was the same as the password I thought I entered. This is not as dangerous
as it seems, though, because if I had made any typographical errors, the password I
2
26
Hour 2
entered the second time (when the system said Retype new passwd:) wouldn’t have
matched the first. In a no-match situation, the system would have warned me that
the information I supplied was inconsistent:
% passwd
Changing password for taylor.
Old password:
New passwd:
Retype new passwd:
Mismatch - password unchanged.
%
Once you change the password, don’t forget it. To reset it to a known value if you
don’t know the current password requires the assistance of a system administrator or
other operator. Renumbering your password can be a catch-22, though: you don’t want to
write down the password because that reduces its secrecy, but you don’t want to forget it,
either. You want to be sure that you pick a good password, too, as described in Task 2.3.
Task 2.3: Picking a Secure Password
If you’re an aficionado of old movies, you are familiar with the thrillers in which the
hoods break into an office and spin the dial on the safe a few times, snicker a bit about
how the boss shouldn’t have chosen his daughter’s birthday as the combination, and crank
open the safe. (If you’re really familiar with the genre, you recall films in which the criminals
rifle through the desk drawers and find the combination of the safe taped to the underside
of a drawer as a fail-safe—or a failed safe, as the case may be.) The moral is that you always
should choose good secret passwords or combinations and keep them secure.
For computers, security is tougher because, in less than an hour, a fast computer system can
test all the words in an English dictionary against your account password. If your password
is kitten or, worse yet, your account name, any semi-competent bad guy could be in your
account and messing with your files in no time.
Many of the more modern UNIX systems have some heuristics, or smarts, built in to the
passwd command; the heuristics check to determine whether what you’ve entered is
reasonably secure.
The tests performed typically answer these questions:
■ Is the proposed password at least six characters long? (A longer password is more
secure.)
■ Does it have both digits and letters? (A mix of both is better.)
■ Does it mix upper- and lowercase letters? (A mix is better.)
■ Is it in the online dictionary? (You should avoid common words.)
■ Is it a name or word associated with the account? (Dave would be a bad password
for my account taylor because my full name on the system is Dave Taylor.)
2
27
Some versions of the passwd program are more sophisticated, and some less, but generally
these questions offer a good guideline for picking a secure password.
1. An easy way to choose memorable and secure passwords is to think of them as
small sentences rather than as a single word with some characters surrounding it. If
you’re a fan of Alexander Dumas and The Three Musketeers, then “All for one and
one for all!” is a familiar cry, but it’s also the basis for a couple of great passwords.
Easily remembered derivations might be all4one or one4all.
2. If you’ve been in the service, you might have the U.S. Army jingle stuck in your
head: “Be All You Can Be” would make a great password, ballucanb. You might
have a self-referential password: account4me or MySekrit would work. If you’re exVice President Dan Quayle, 1Potatoe could be a memorable choice (potatoe by
itself wouldn’t be particularly secure because it lacks digits and lacks uppercase
letters, and because it’s a simple variation on a word in the online dictionary).
3. Another way to choose passwords is to find acronyms that have special meaning to
you. Don’t choose simple ones—remember, short ones aren’t going to be secure.
But, if you have always heard that “Real programmers don’t eat quiche!” then
Rpdeq! could be a complex password that you’ll easily remember.
4. Many systems you use every day require numeric passwords to verify your identity,
including the automated-teller machine (with its PIN number), government
agencies (with the Social Security number), and the Department of Motor Vehicles
(your driver’s license number or vehicle license). Each of these actually is a poor
UNIX password: it’s too easy for someone to find out your license number or
Social Security number.
JUST A MINUTE
The important thing is that you come up with a strategy of your own for
choosing a password that is both memorable and secure. Then, keep the
password in your head rather than write it down.
Why be so paranoid? For a small UNIX system that will sit on your desk in your office
and won’t have any other users, a high level of concern for security is, to be honest,
unnecessary. As with driving a car, though, it’s never too early to learn good habits. Any
system that has dial-up access or direct-computer-network access—you might need to use
such a system—is a likely target for delinquents who relish the intellectual challenge of
breaking into an account and altering and destroying files and programs purely for
amusement.
2
Hour 2
28
The best way to avoid trouble is to develop good security habits now when you’re first
learning about UNIX—learn how to recognize what makes a good, secure password; pick one
for your account; and keep it a secret.
If you ever need to let someone else use your account for a short time, remember that you
can use the passwd command to change your secure password to something less secure. Then,
you can let that person use the account, and, when he or she is done, you can change the
password back to your original password.
With that in mind, log in again to your UNIX system now, and try changing your password.
First, change it to easy and see if the program warns you that easy is too short or otherwise
a poor choice. Then, try entering two different secret passwords to see if the program notices
the difference. Finally, pick a good password, using the preceding guidelines and suggestions,
and change your account password to be more secure.
Task 2.4: Who Are You?
While you’re logged in to the system, you can learn a few more UNIX commands,
including a couple that can answer a philosophical conundrum that has bothered
men and women of thought for thousands of years: Who am I?
1. The easiest way to find out “who you are” is to enter the whoami command:
% whoami
taylor
%
Try it on your system. The command lists the account name associated with the
current login.
2. Ninety-nine percent of the commands you type with UNIX don’t change if you
modify the punctuation and spacing. With whoami, however, adding spaces to
transform the statement into proper English—that is, entering who am i—dramatically changes the result. On my system, I get the following results:
% who am i
mentor.utech.edu!taylor
%
ttyp4
Feb 8 14:34
This tells me quite a bit about my identity on the computer, including the name of
the computer itself, my account name, and where and when I logged in. Try the
command on your system and see what results you get.
In this example, mentor is a hostname—the name of the computer I am logged in
to—and utech.edu is the full domain name—the address of mentor. The exclamation point (!) separates the domain name from my account name, taylor. The
2
29
(pronounced “tee-tee-why-pea-four”) is the current communication line I’m
using to access mentor, and 5 October at 2:34pm is when I logged in to mentor
today.
ttyp4
JUST A MINUTE
UNIX is full of oddities that are based on historical precedent. One is “tty”
to describe a computer or terminal line. This comes from the earliest UNIX
systems in which Digital Equipment Corporation teletypewriters would be
hooked up as interactive devices. The teletypewriters quickly received the
nickname “tty,” and all these years later, when people wouldn’t dream of
hooking up a teletypewriter, the line is still known as a tty line.
3. One of the most dramatic influences UNIX systems have had on the computing
community is the propensity for users to work together on a network, hooked up
by telephone lines and modems (the predominant method until the middle to late
1980s) or by high-speed network connections to the Internet (a more common
type of connection today). Regardless of the connection, however, you can see that
each computer needs a unique identifier to distinguish it from others on the
network. In the early days of UNIX, systems had unique hostnames, but as
hundreds of systems have grown into the tens-of-thousands, that proved to be an
unworkable solution.
4. The alternative was what’s called a “domain-based naming scheme,” where systems
are assigned unique names within specific subsets of the overall network. Consider
the output that was shown in instruction 2, for example:
mentor.utech.edu!taylor
ttyp4
Feb 11 14:34
The computer I use is within the .edu domain (read the hostname and domain—
mentor.utech.edu —from right to left), meaning that the computer is located at an
educational institute. Then, within the educational institute subset of the network,
utech is a unique descriptor, and, therefore, if other UTech universities existed,
they couldn’t use the same top-level domain name. Finally, mentor is the name of
the computer itself.
5. Like learning to read addresses on envelopes, learning how to read domain names
can unlock a lot of information about a computer and its location. For example,
lib.stanford.edu is the library computer at Stanford University, and
ccgate.infoworld.com tells you that the computer is at InfoWorld, a commercial
computer site, and that its hostname is ccgate. You learn more about this a few
hours down the road when you learn how to use electronic mail to communicate
with people throughout the Internet.
2
Hour 2
30
6. Another way to find out who you are in UNIX is the id command. The purpose of
this command is to tell you what group or groups you’re in and the numeric
identifier for your account name (known as your user ID number or user ID). Enter
id and see what you get. I get the following result:
% id
uid=211(taylor)
%
JUST A MINUTE
gid=50(users0) groups=50(users0)
If you enter id, and the computer returns a different result or indicates that
you need to specify a filename, don’t panic. On many Berkeley-derived
systems, the id command is used to obtain low-level information about
files.
7. In this example, you can see that my account name is taylor and that the numeric
equivalent, the user ID, is 211. (Here it’s abbreviated as uid—pronounce it “youeye-dee” to sound like a UNIX expert.) Just as the account name is unique on a
system, so also is the user ID. Fortunately, you rarely, if ever, need to know these
numbers, so focus on the account name and group name.
8. Next, you can see that my group ID (or gid) is 50, and that group number 50 is
known as the users0 group. Finally, users0 is the only group to which I belong.
On another system, I am a member of two different groups:
% id
uid=103(taylor) gid=10(staff) groups=10(staff),44(ftp)
%
Although I have the same account name on this system (taylor), you can see that
my user ID and group ID are both different from the earlier example. Note also
that I’m a member of two groups: the staff group, with a group ID of 10, and the
ftp group, with a group ID of 44.
Later, you learn how to set protection modes on your files so that people in your group
can read your files, but those not in your group are barred from access. You’ve now
learned a couple of different ways to have UNIX give you some information about your
account.
Task 2.5: Finding Out What Other Users Are
Logged in to the System
The next philosophical puzzle that you can solve with UNIX is “Who else is there?”
The answer, however, is rather restricted, limited to only those people currently
2
31
logged in to the computer at the same time. Three commands are available to get you this
information, based on how much you’d like to learn about the other users: users, who,
and w.
1. The simplest of the commands is the users command, which lists the account
names of all people using the system:
% users
david mark taylor
%
2. In this example, david and mark are also logged in to the system with me. Try this
on your computer and see what other users—if any—are logged in to your computer system.
3. A command that you’ve encountered earlier in this hour can be used to find out
who is logged on to the system, what line they’re on, and how long they’ve been
logged in. That command is who:
% who
taylor
david
mark
%
ttyp0
ttyp2
ttyp4
Oct
Oct
Oct
8 14:10
4 09:08
8 12:09
(limbo)
(calliope)
(dent)
Here, you can see that three people are logged in, taylor (me), david, and mark.
Further, you can now see that david is logged in by connection ttyp2 and has been
connected since October 4 at 9:08 a.m. He is connected from a system called
calliope. You can see that mark has been connected since just after noon on
October 8 on line ttyp4 and is coming from a computer called dent. Note that I
have been logged in since 14:10, which is 24-hour time for 2:10 p.m. UNIX
doesn’t always indicate a.m. or p.m.
The user and who commands can inform you who is using the system at any particular
moment, but how do you find out what they’re doing?
Task 2.6: What Is Everyone Doing on the
Computer?
To find out what everyone else is doing, there’s a third command, w, that serves as
a combination of “Who are they?” and “What are they doing?”
2
Hour 2
32
1. Consider the following output from the w command:
% w
2:12pm
User
taylor
david
mark
%
up 7 days, 5:28,
tty
login@
ttyp0
2:10pm
ttyp2
Mon 9am
ttyp4
12:09pm
3 users, load average: 0.33, 0.33, 0.02
idle
JCPU
PCPU what
2
w
2:11
2:04
1:13 xfax
2:03
-csh
This is a much more complex command, offering more information than either
users or who. Notice that the output is broken into different areas. The first line
summarizes the status of the system and, rather cryptically, the number of programs that the computer is running at one time. Finally, for each user, the output
indicates the user name, the tty, when the user logged in to the system, how long
it’s been since the user has done anything (in minutes and seconds), the combined
CPU time of all jobs the user has run, and the amount of CPU time taken by the
current job. The last field tells you what you wanted to know in the first place:
what are the users doing?
In this example, the current time is 2:12 p.m., and the system has been up for 7
days, 5 hours, and 28 minutes. Currently 3 users are logged in, and the system is
very quiet, with an average of 0.33 jobs submitted (or programs started) in the last
minute; 0.33, on average, in the last 5 minutes; and 0.02 jobs in the last 15
minutes.
User taylor is the only user actively using the computer (that is, who has no idle
time) and is using the w command. User david is running a program called xfax,
which has gone for quite a while without any input from the user (2 hours and 11
minutes of idle time). The program already has used 1 minute and 13 seconds of
CPU time, and overall, david has used over 2 minutes of CPU time. User mark has
a C shell running, -csh. (The leading dash indicates that this is the program that
the computer launched automatically when mark logged in. This is akin to how the
system automatically launched the Finder on a Macintosh on startup.) User mark
hasn’t actually done anything yet: notice there is no accumulated computer time
for that account.
2. Now it’s your turn. Try the w command on your system and see what kind of
output you get. Try to interpret all the information based on the explanation here.
One thing is certain: your account should have the w command listed as what
you’re doing.
2
33
On a multiuser UNIX system, the w command gives you a quick and easy way to see
what’s going on.
Task 2.7: Checking the Current Date and Time
You’ve learned how to orient yourself on a UNIX system, and you are able now to
figure out who you are, who else is on the system, and what everyone is doing. What
about the current time and date?
2
1. Logic suggests that time shows the current time, and date the current date; but this
is UNIX, and logic doesn’t always apply. In fact, consider what happens when I
enter time on my system:
% time
14.5u 17.0s 29:13 1% 172+217io 160pf+1w
%
The output is cryptic to the extreme and definitely not what you’re interested in
finding out. Instead, the program is showing how much user time, system time,
and CPU time has been used by the command interpreter itself, broken down by
input/output operations and more. This is not something I’ve ever used in 15 years
of working with UNIX.
2. Well, time didn’t work, so what about date?
% date
Tue Oct 5 15:03:41 EST 1993
%
That’s more like it!
3. Try the date command on your computer and see if the output agrees with your
watch.
How do you think date keeps track of the time and date when you’ve turned the
computer off ? Does the computer know the correct time if you unplug it for a
few hours? (I hope so. Almost all computers today have little batteries inside for just this
purpose.)
Task 2.8: Looking at a Calendar
Another useful utility in UNIX is the cal command, which shows a simple calendar
for the month or year specified.
Hour 2
34
1. To confirm that 5 October 1993 is a Tuesday, turn to your computer and enter
cal 10 93. You should see the following:
% cal 10 93
October 93
S M Tu W Th F S
1 2 3 4 5
6 7 8 9 10 11 12
13 14 15 16 17 18 19
20 21 22 23 24 25 26
27 28 29 30 31
%
2. If you look closely, you’ll find that there’s a bit of a problem here. October 5 is
shown as a Saturday rather than a Tuesday as expected.
The reason is that cal can list any year from A.D. 0. In fact, what you have on
your screen is how the month of October would have looked in A.D. 93, 1900
years ago.
JUST A MINUTE
This is a bit misleading because Western society uses the Julian calendar,
adopted in 1752. Before that, the program should really list Gregorianformat monthly calendars, but it can’t, so don’t use this as a historical
reference for ascertaining what day of the week the Emperor Hadrian was
born.
3. To find out the information that you want, you’ll need to specify to the cal
program both the month and full year:
% cal 10 1993
October 1993
S M Tu W Th F S
1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30
31
%
This is correct. The 5th of October in 1993 is indeed a Tuesday. On some systems,
cal has no intelligent default action, so entering cal doesn’t simply list the monthly
calendar for the current month. Later you’ll learn how to write a simple shell script
to do just that. For now, turn to your system and enter cal to see what happens.
2
35
4. My favorite example of the cal program is to ask for the year 1752, the year when
the Western calendar switched from Gregorian to Julian. Note particularly the
month of September, during which the switch actually occurred.
% cal 1752
S
M Tu
5 6 7
12 13 14
19 20 21
26 27 28
S
M Tu
5 6 7
12 13 14
19 20 21
26 27 28
S
M Tu
5 6 7
12 13 14
19 20 21
26 27 28
Jan
W Th
1 2
8 9
15 16
22 23
29 30
Apr
W Th
1 2
8 9
15 16
22 23
29 30
Jul
W Th
1 2
8 9
15 16
22 23
29 30
F S
3 4
10 11
17 18
24 25
31
F S
3 4
10 11
17 18
24 25
F S
3 4
10 11
17 18
24 25
31
Oct
S M Tu W Th F S
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28
29 30 31
S
1752
Feb
M Tu W Th
F
S
1
8
15
22
29
2 3 4 5 6 7
9 10 11 12 13 14
16 17 18 19 20 21
23 24 25 26 27 28
May
S M Tu W Th F S
1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30
31
Aug
S M Tu W Th F S
1
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
30 31
Nov
S M Tu W Th F S
1 2 3 4
5 6 7 8 9 10 11
12 13 14 15 16 17 18
19 20 21 22 23 24 25
26 27 28 29 30
Mar
S M Tu W Th F
1 2 3 4 5 6
8 9 10 11 12 13
15 16 17 18 19 20
22 23 24 25 26 27
29 30 31
Jun
S M Tu W Th F
1 2 3 4 5
7 8 9 10 11 12
14 15 16 17 18 19
21 22 23 24 25 26
28 29 30
S
7
14
21
28
S
6
13
20
27
Sep
M Tu W Th F S
1 2 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30
S
S
M Tu
Dec
W Th
3 4 5 6 7
10 11 12 13 14
17 18 19 20 21
24 25 26 27 28
31
F S
1 2
8 9
15 16
22 23
29 30
%
You can experiment with cal and easily find out fun information—for example, what
day of the week you or your parents were born. If you’re curious about whether
Christmas 1999 is on a weekend, cal can answer that question, too.
When you used cal, you entered the name of the command and then some additional
information to indicate the exact action you desired. You tried both cal 10 93 and cal 10
1993. In UNIX parlance, the first word is the command, and the subsequent words are
arguments or options to the command. A special class of options are those that begin with a
single dash, called flags, and you’ll learn about those starting in the next hour.
2
Hour 2
36
Simple Math with UNIX
Having both an internal wall clock and an internal calendar, UNIX seems to have much of
what you need in an office. One piece that’s missing now, however, is a simple desktop
calculator. UNIX offers two different types of calculator, although neither rightly can be
called simple.
Mathematicians talk about infix and postfix notation as two different ways to write an
expression, the former having the operation embedded in the operators, and the latter having
all the operators listed, followed by the operation required. Table 2.1 lists some examples of
a mathematical expression in both formats.
Table 2.1. Comparing infix and postfix notation.
Infix
Postfix
75*0.85
75
0.85
(37*1.334)+44
37
1.334
cos(3.45)/4
3.45
cos
*
*
44
4
/
+
You’re probably familiar with the infix notation, which is the form used in math textbooks
throughout the world. Lots of calculators can work this way, too; you’d press the keys
1 + 1 = to find out that 1 plus 1 equals 2.
Some calculators offer the postfix alternative, also known as (reverse) Polish notation,
invented by Polish mathematician and logician Jan Lukasiewicz. Notably, for many years
Hewlett-Packard has been making calculators that work with RPN notation. On an HP
calculator, you’d press the keys 1 Enter 1 + to find out that 1 plus 1 equals 2.
Notice that, although parentheses were required in the second equation in the table when
using infix notation, parentheses weren’t necessary to force a specific order of evaluation with
postfix. Remember that in math you always work from the inside of the parentheses outward,
so (3 * 4) + 8 is solved by multiplying 3 by 4, then adding 8, and that process is exactly what
RPN mimics.
UNIX offers two calculator programs, one with infix notation and one with postfix notation.
Task 2.9: Using the bc Infix Calculator
The first calculator to learn is bc, the UNIX infix-notation calculator.
2
37
1. To use the infix calculator, enter the following command:
% bc
Nothing happens—no prompt, nothing. The reason is that bc, like its RPN cousin
dc, waits for you to enter commands. The quit command lets you leave the
program. You can see how it works by seeing how I solve the first and second
mathematical equations of Table 2.1:
% bc
75 * 0.85
63.75
(37*1.334)+44
93.358
quit
%
2. Unfortunately, bc is, in many ways, a typical UNIX command. Consider what
happens when I enter help, hoping for some clue on how to use the bc program:
% bc
help
syntax error on line 1, teletype
3. This is not very helpful. If you get stuck in a command, there are two surefire ways
to escape. Control-d (holding down the Control—also called Ctrl—key on your
keyboard and simultaneously pressing the d key) indicates that you have no further
input, which often causes programs to quit. If that fails, Control-c kills the program, that is, forces it to quit immediately.
The bc command has a number of powerful and useful options, as shown in
Table 2.2.
Table 2.2. Helpful bc commands.
Notation
Description of Function
sqrt(n)
Square root of n
Remainder
To the power of (3^5 is 3 to the power of 5)
Sine(n)
Cosine(n)
Exponential(n)
Log(n)
%
^
s(n)
c(n)
e(n)
l(n)
2
Hour 2
38
4. If you wanted to calculate the sine of 4.5243 to the third power, you could do it
with bc. You need to be sure, however, that the system knows you’re working with
higher math functions by specifying the command flag -l math (or, in some cases,
just -l):
% bc -l math
s ( 4.5243 ^ 3 )
-.99770433540886100879
quit
%
If you try this on your calculator, you probably won’t get a result quite as precise as
this. The bc and dc commands both work with extended precision, allowing for
highly accurate results.
Task 2.10: Using the dc Postfix Calculator
By contrast, the dc command works with the postfix notation, and each number or
operation must be on its own line. Further, the result of an operation isn’t
automatically shown; you have to enter p to see the most recently calculated result.
1. To use dc for the calculations shown previously, enter the following characters
shown in bold. The result follows each completed entry.
% dc
75
0.85
*
p
63.75
37
1.334
*
44
+
p
93.358
quit
%
2. The set of commands available in dc are different because dc addresses a different
set of mathematical equations. The dc command is particularly useful if you need
to work in a non-decimal base. (For example, some older computer systems worked
in octal, a base-8 numbering system. The number 210 in octal, therefore, represents 2 * 8 * 8 + 1 * 8 + 0, or 136 in decimal.) Table 2.3 summarizes some of the
most useful commands available in dc.
2
39
Table 2.3. Helpful commands in dc.
Notation
Description of Function
v
Square root
Set radix (numeric base) for input
Set radix for output
Print top of stack
All values in the stack are printed
i
o
p
f
3. For example, I used dc to verify that 210 (octal) is indeed equal to 136 (decimal):
% dc
8
i
210
p
136
With a little work, you can use different numeric bases within the bc program, so
unless you’re really used to the RPN notation, it’s probably best to remember the bc
command when you think of doing some quick calculations in UNIX.
JUST A MINUTE
I find both bc and dc ridiculously difficult to use, so I keep a small handheld calculator by my computer. For just about any task, simply using the
calculator is faster than remembering the notational oddities of either
UNIX calculator program. Your mileage may vary, of course.
If you run the X Window System, the UNIX graphical interface, there are
several calculator programs that look exactly like a hand-held calculator.
If you’re old enough, you’ll remember the early 1980s as the time when IBM introduced the
PC and the industry was going wild, predicting that within a few years every home would have
a PC and that everyone would use PCs for balancing checkbooks and keeping track of recipes.
Fifteen years later, few people in fact use computers as part of their cooking ritual, although
checkbook balancing programs are amazingly popular. The point is that some tasks can be
done by computer but are sometimes best accomplished through more traditional means. If
you have a calculator and are comfortable using it, the calculator is probably a better solution
than learning how to work with bc to add a few numbers.
There are definitely situations where having the computer add the numbers for you is quite
beneficial—particularly when there are a lot of them—but if you’re like me, you rarely
encounter that situation.
2
Hour 2
40
Summary
This hour focused on giving you the skills required to log in to a UNIX system, figure out
who you are and what groups you’re in, change your password, and log out again. You also
learned how to list the other users of the system, find out what UNIX commands they’re
using, check the date and time, and even show a calendar view of almost any month or year
in history. Finally, you learned some of the power of two similar UNIX utilities, bc and dc,
the two UNIX desktop calculators.
Workshop
in the next hour.
Key Terms
account name This is the official one-word name by which the UNIX system knows you:
mine is taylor. (See also account in Hour 1.)
domain name UNIX systems on the Internet, or any other network, are assigned a domain
within which they exist. This is typically the company (for example, sun.com for Sun
Microsystems) or institution (for example, lsu.edu for Louisiana State University). The
domain name is always the entire host address, except the host name itself. (See also host
name.)
flags Arguments given to a UNIX command that are intended to alter its behavior are
called flags. They’re always prefaced by a single dash. As an example, the command line
ls -l /tmp has ls as the command itself, -l as the flag to the command, and /tmp as the
argument.
heuristic A set of well-defined steps or a procedure for accomplishing a specific task.
host name UNIX computers all have unique names assigned by the local administration
team. The computers I use are limbo, well, netcom, and mentor, for example. Enter hostname
to see what your system is called.
login A synonym for account name, this also can refer to the actual process of connecting
to the UNIX system and entering your account name and password to your account.
user ID A synonym for account name.
2
41
Questions
1. Why can’t you have the same account name as another user? How about user ID?
Can you have the same uid as someone else on the system?
2. Which of the following are good passwords, based on the guidelines you’ve learned
in this hour?
foobar
4myMUM
Blk&Blu
234334
Laurie
Hi!
2cool.
rolyat
j j kim
3. Are the results of the two commands who am i and whoami different? If so, explain
how. Which do you think you’d rather use when you’re on a new computer?
4. List the three UNIX commands to find out who is logged on to the system. Talk
about the differences between the commands.
5. One of the commands in the answer to question 4 indicates how long the system
has been running (in the example, it’d been running for seven days). What value do
you think there is for keeping track of this information?
6. If you can figure out what other people are doing on the computer, they can figure
out what you’re doing, too. Does that bother you?
7. What day of the week were you born? What day of the week is July 4, 1997? For
that matter, what day of the week was July 4, 1776?
8. Solve the following mathematical equations using both dc and bc, and then explain
which command you prefer.
454 * 3.84
sin(3.1415)
log(2.45)+log(3)
2^16
The next hour focuses on the UNIX hierarchical file system. You learn about how the system
is organized, how it differs from Macintosh and DOS hierarchical file systems, the difference
between “relative” and “absolute” filenames, and what the mysterious “.” and “..” directories
are. You also learn about the env, pwd, and cd commands, and the HOME and PATH environment
variables.
2
Moving About in the File System
Hour
3
Moving About the File
System
This third hour focuses on the UNIX hierarchical file system. You learn about
how the system is organized, how it differs from the Macintosh and DOS
hierarchical file systems, the difference between “relative” and “absolute”
filenames, and what the mysterious “.” and “..” directories are. You also learn
about the env, pwd, and cd commands and the HOME and PATH environment
variables.
Goals for This Hour
In this hour, you learn
■
■
■
■
■
■
What a hierarchical file system is all about
How the UNIX file system is organized
How Mac and PC file systems differ from UNIX
The difference between relative and absolute filenames
About hidden files in UNIX
About the special directories “.” and “..”
43
3
Hour 3
44
■
■
■
■
The env command
About user environment variables, PATH and HOME
How to find where you are with pwd
How to move to another location with cd
The previous hour introduced a plethora of UNIX commands, but this hour takes a more
theoretical approach, focusing on the UNIX file system, how it’s organized, and how you can
navigate it. This hour focuses on the environment that tags along with you as you move about,
particularly the HOME and PATH variables. After that is explained, you learn about the env
command as an easy way to show environment variables, and you learn the pwd and cd pair
of commands for moving about directly.
What a Hierarchical File System
Is All About
In a nutshell, a hierarchy is a system organized by graded categorization. A familiar example
is the organizational structure of a company, where workers report to supervisors and
supervisors report to middle managers. Middle managers, in turn, report to senior managers,
and senior managers report to vice-presidents, who report to the president of the company.
Graphically, this hierarchy looks like Figure 3.1.
Figure 3.1.
A typical organizational
hierarchy.
You’ve doubtless seen this type of illustration before, and you know that a higher position
indicates more control. Each position is controlled by the next highest position or row. The
president is top dog of the organization, but each subsequent manager is also in control of
his or her own small fiefdom.
To understand how a file system can have a similar organization, simply imagine each of the
managers in the illustration as a “file folder” and each of the employees as a piece of paper,
filed in a particular folder. Open any file cabinet, and you probably see things organized this
3
45
way: filed papers are placed in labeled folders, and often these folders are filed in groups under
specific topics. The drawer might then have a specific label to distinguish it from other
drawers in the cabinet, and so on.
That’s exactly what a hierarchical file system is all about. You want to have your files located
in the most appropriate place in the file system, whether at the very top, in a folder, or in a
nested series of folders. With careful usage, a hierarchical file system can contain hundreds
or thousands of files and still allow users to find any individual file quickly.
On my computer, the chapters of this book are organized in a hierarchical fashion, as shown
in Figure 3.2.
Figure 3.2.
File organization for the
chapters of Teach
Yourself UNIX in 24
Hours.
3
Task 3.1: The UNIX File System Organization
A key concept enabling the UNIX hierarchical file system to be so effective is that
anything that is not a folder is a file. Programs are files in UNIX, device drivers are
files, documents and spreadsheets are files, your keyboard is represented as a file, your display
is a file, and even your tty line and mouse are files.
What this means is that as UNIX has developed, it has avoided becoming an ungainly mess.
UNIX does not have hundreds of cryptic files stuck at the top (this is still a problem in DOS)
or tucked away in confusing folders within the System Folder (as with the Macintosh).
The top level of the UNIX file structure (/) is known as the root directory or slash directory,
and it always has a certain set of subdirectories, including bin, dev, etc, lib, mnt, tmp, and usr.
There can be a lot more, however. Listing 3.1 shows files found at the top level of the mentor
file system (the system I work on). Typical UNIX directories are shown followed by a slash
in the listing.
AA
OLD/
archive/
ats/
backup/
bin/
boot
core
dev/
diag/
dynix
etc/
flags/
gendynix
lib/
lost+found/
mnt/
net/
rf/
stand/
sys/
tftpboot/
tmp/
usera/
userb/
userc/
users/
usere/
users/
usr/
var/
Hour 3
46
You can obtain a listing of the files and directories in your own top-level directory by using
the ls –C -F / command. (You’ll learn all about the ls command in the next hour. For now,
just be sure that you enter exactly what’s shown in the example.)
On a different computer system, here’s what I see when I enter that command:
% ls –C -F /
Mail/
export/
News/
home/
add_swap/
kadb*
apps/
layout
archives/
lib@
bin@
lost+found/
boot
mnt/
cdrom/
net/
chess/
news/
dev/
nntpserver
etc/
pcfs/
public/
reviews/
sbin/
sys@
tftpboot/
tmp/
usr/
utilities/
var/
vmunix*
In this example, any filename that ends with a slash (/) is a folder (UNIX calls these
directories). Any filename that ends with an asterisk (*) is a program. Anything ending with
an at sign (@) is a symbolic link, and everything else is a normal, plain file.
As you can see from these two examples, and as you’ll immediately find when you try the
command yourself, there is much variation in how different UNIX systems organize the toplevel directory. There are some directories and files in common, and once you start examining
the contents of specific directories, you’ll find that hundreds of programs and files always
show up in the same place from UNIX to UNIX.
It’s as if you were working as a file clerk at a new law firm. Although this firm might have a
specific approach to filing information, the approach may be similar to the filing system of
other firms where you have worked in the past. If you know the underlying organization, you
can quickly pick up the specifics of a particular organization.
Try the command ls –C -F / on your computer system, and identify, as previously
explained, each of the directories in your resultant listing.
The output of the ls command shows the files and directories in the top level of your
system. Next, you learn what they are.
The bin Directory
In UNIX parlance, programs are considered executables because users can execute them. (In
this case, execute is a synonym for run, not an indication that you get to wander about
murdering innocent applications!) When the program has been compiled (usually from a C
listing), it is translated into what’s called a binary format. Add the two together, and you have
a common UNIX description for an application—an executable binary.
3
47
It’s no surprise that the original UNIX developers decided to have a directory labeled
“binaries” to store all the executable programs on the system. Remember the primitive
teletypewriter discussed in the last hour? Having a slow system to talk with the computer had
many ramifications that you might not expect. The single most obvious one was that
everything became quite concise. There were no lengthy words like binaries or listfiles,
but rather succinct abbreviations: bin and ls are, respectively, the UNIX equivalents.
The bin directory is where all the executable binaries were kept in early UNIX. Over time,
as more and more executables were added to UNIX, having all the executables in one place
proved unmanageable, and the bin directory split into multiple parts (/bin, /sbin, /usr/bin).
The dev Directory
Among the most important portions of any computer are its device drivers. Without them,
you wouldn’t have any information on your screen (the information arrives courtesy of the
display device driver). You wouldn’t be able to enter information (the information is read and
given to the system by the keyboard device driver), and you wouldn’t be able to use your
floppy disk drive (managed by the floppy device driver).
Earlier, you learned how almost anything in UNIX is considered a file in the file system, and
the dev directory is an example. All device drivers—often numbering into the hundreds—
are stored as separate files in the standard UNIX dev (devices) directory. Pronounce this
directory name “dev,” not “dee-ee-vee.”
The etc Directory
UNIX administration can be quite complex, involving management of user accounts, the file
system, security, device drivers, hardware configurations, and more. To help, UNIX
designates the etc directory as the storage place for all administrative files and information.
Pronounce the directory name either “ee-tea-sea”, “et-sea,” or “etcetera.” All three pronunciations are common.
The lib Directory
Like your neighborhood community, UNIX has a central storage place for function and
procedural libraries. These specific executables are included with specific programs, allowing
programs to offer features and capabilities otherwise unavailable. The idea is that if programs
want to include certain features, they can reference just the shared copy of that utility in the
UNIX library rather than having a new, unique copy.
In the previous hour, when you were exploring the dc calculator, you used the command
dc -l math to access trigonometric functions. The -l math was to let dc know that you wanted
to include the functions available through the math library, stored in the lib directory.
3
Hour 3
48
Many of the more recent UNIX systems also support what’s called dynamic linking, where
the library of functions is included on-the-fly as you start up the program. The wrinkle is that
instead of the library reference being resolved when the program is created, it’s resolved only
when you actually run the program itself.
Pronounce the directory name “libe” or “lib” (to rhyme with the word bib).
The lost+found Directory
With multiple users running many different programs simultaneously, it’s been a challenge
over the years to develop a file system that can remain synchronized with the activity of the
computer. Various parts of the UNIX kernel—the brains of the system—help with this
problem. When files are recovered after any sort of problem or failure, they are placed here,
in the lost+found directory, if the kernel cannot ascertain the proper location in the file
system. This directory should be empty almost all the time.
This directory is commonly pronounced “lost and found” rather than “lost plus found.”
The mnt and sys Directories
The mnt (pronounced “em-en-tea”) and sys (pronounced “sis”) directories also are safely
ignored by UNIX users. The mnt directory is intended to be a common place to mount
external media—hard disks, removable cartridge drives, and so on—in UNIX. On many
systems, though not all, sys contains files indicating the system configuration.
The tmp Directory
A directory that you can’t ignore, the tmp directory—say “temp”—is used by many of the
programs in UNIX as a temporary file-storage space. If you’re editing a file, for example, the
program makes a copy of the file and saves it in tmp, and you work directly with that, saving
the new file back to your original file only when you’ve completed your work.
On most systems, tmp ends up littered with various files and executables left by programs that
don’t remove their own temporary files. On one system I use, it’s not uncommon to find
10–30 megabytes of files wasting space here.
Even so, if you’re manipulating files or working with copies of files, tmp is the best place to
keep the temporary copies of files. Indeed, on some UNIX workstations, tmp actually can be
the fastest device on the computer, allowing for dramatic performance improvements over
working with files directly in your home directory.
The usr Directory
Finally, the last of the standard directories at the top level of the UNIX file system hierarchy
is the usr—pronounced “user”—directory. Originally, this directory was intended to be the
central storage place for all user-related commands. Today, however, many companies have
their own interpretation, and there’s no telling what you’ll find in this directory.
3
49
Standard practice is that /usr contains UNIX operating system binaries.
JUST A MINUTE
Other Miscellaneous Stuff at the Top Level
Besides all the directories previously listed, a number of other directories and files commonly
occur in UNIX systems. Some files might have slight variations in name on your computer,
so when you compare your listing to the following files and directories, be alert for possible
alternative spellings.
A file you must have to bring up UNIX at all is one usually called unix or vmunix, or named
after the specific version of UNIX on the computer. The file contains the actual UNIX
operating system. The file must have a specific name and must be found at the top level of
the file system. Hand-in-hand with the operating system is another file called boot, which
helps during initial startup of the hardware.
Notice on one of the previous listings that the files boot and dynix appear. (DYNIX is the
name of the particular variant of UNIX used on Sequent computers.) By comparison, the
listing from the Sun Microsystems workstation shows boot and vmunix as the two files.
Another directory that you might find in your own top-level listing is diag—pronounced
“dye-ag”—which acts as a storehouse for diagnostic and maintenance programs. If you have
any programs within this directory, it’s best not to try them out without proper training!
The home directory, also sometimes called users, is a central place for organizing all files unique
to a specific user. Listing this directory is usually an easy way to find out what accounts are
on the system, too, because by convention each individual account directory is named after
the user’s account name. On one system I use, my account is taylor, and my individual
account directory is also called taylor. Home directories are always created by the system
administrator.
The net directory, if set up correctly, is a handy shortcut for accessing other computers on
your network.
The tftpboot directory is a relatively new feature of UNIX. The letters stand for “trivial file
transfer protocol boot.” Don’t let the name confuse you, though; this directory contains
versions of the kernel suitable for X Window System-based terminals and diskless workstations to run UNIX.
Some UNIX systems have directories named for specific types of peripherals that can be
attached. On the Sun workstation, you can see examples with the directories cdrom and pcfs.
The former is for a CD-ROM drive and the latter for DOS-format floppy disks.
There are many more directories in UNIX, but this will give you an idea of how things are
organized.
3
50
Hour 3
How Mac and PC File Systems Differ from
the UNIX File System
Although the specific information is certainly different, some parallels do exist between the
hierarchical file system structures of the UNIX and the Macintosh systems.
For example, on the Macintosh, folders are distinguished by their icons. The common folders
you’ll find on all Macs include System Folder and Trash. Within the system folder, all Macs
have a variety of system-related files, including Finder, System, and Clipboard. Folders
include Extensions, Preferences, and Control Panels.
By comparison, DOS requires few files be present for the system to be usable: command.com
must be present, and autoexec.bat and config.sys usually are present. Most DOS systems
have all the commands neatly tucked into the \DOS directory on the system, but sometimes
these commands appear at the very top level.
Directory Separator Characters
If you look at the organizational chart presented earlier in this hour, you can see that
employees are identified simply as “employee” where possible. Because each has a unique
path upwards to the president, each has a unique identifier if all components of the path
upward are specified.
For example, the rightmost of the four employees could be described as “Employee managed
by Jr. Manager 4, managed by Senior Manager 3, managed by Vice-President 2, managed
by the President.” Using a single character, instead of “managed by,” can considerably
shorten the description: Employee/Jr. Manager 4/Senior Manager 3/Vice-President 2/
President. Now consider the same path specified from the very top of the organization
downward: President/Vice-President 2/Senior Manager 3/Jr. Manager 4/Employee.
Because only one person is at the top, that person can be safely dropped from the path without
losing the uniqueness of the descriptor: /Vice-President 2/Senior Manager 3/Jr. Manager 4/
Employee.
In this example, the / (pronounce it “slash”) is serving as a directory separator character, a
convenient shorthand to indicate different directories in a path.
The idea of using a single character isn’t unique to UNIX, but using the slash is unusual. On
the Macintosh, the system uses a colon to separate directories in a pathname. (Next time
you’re on a Mac, try saving a file called test:file and see what happens.) DOS uses a
backslash: \DOS indicates the DOS directory at the top level of DOS. The characters /tmp
indicate the tmp directory at the top level of the UNIX file system, and :Apps is a folder called
Apps at the top of the Macintosh file system.
3
51
On the Macintosh, you rarely encounter the directory delineator because the system has a
completely graphical interface. Windows also offers a similar level of freedom from having
to worry about much of this complexity, although you’ll still need to remember whether “A:”
is your floppy disk or hard disk drive.
The Difference Between Relative and
Absolute Filenames
Specifying the location of a file in a hierarchy to ensure that the filename is unique is known
in UNIX parlance as specifying its absolute filename. That is, regardless of where you are
within the file system, the absolute filename always specifies a particular file. By contrast,
relative filenames are not unique descriptors.
To understand, consider the files shown in Figure 3.3.
Figure 3.3.
A simple hierarchy
of files.
If you are currently looking at the information in the Indiana directory, Bldgs uniquely
describes one file: the Bldgs file in the Indiana directory. That same name, however, refers
to a different file if you are in the California or Washington directories. Similarly, the
directory Personnel leaves you with three possible choices until you also specify which state
you’re interested in.
As a possible scenario, imagine you’re reading through the Bldgs file for Washington and some
people come into your office, interrupting your work. After a few minutes of talk, they
comment about an entry in the Bldgs file in California. You turn to your UNIX system and
bring up the Bldgs file, and it’s the wrong file. Why? You’re still in the Washington directory.
These problems arise because of the lack of specificity of relative filenames. Relative filenames
describe files that are referenced relative to an assumed position in the file system. In Figure
3.3, even Personnel/Taylor,D. isn’t unique because that can be found in both Indiana and
Washington .
3
Hour 3
52
To avoid these problems, you can apply the technique you learned earlier, specifying all
elements of the directory path from the top down. To look at the Bldgs file for California,
you could simply specify /California/Bldgs. To check the Taylor,D. employee in Indiana,
you’d use /Indiana/Personnel/Taylor,D., which is different, you’ll notice, from the
employee /Washington/Personnel/Taylor,D..
Learning the difference between these two notations is crucial to surviving the complexity of
a hierarchical file system used with UNIX. Without it, you’ll spend half your time verifying
that you are where you think you are, or, worse, not moving about at all, not taking advantage
of the organizational capabilities.
If you’re ever in doubt as to where you are or what file you’re working with in UNIX, simply
specify its absolute filename. You always can differentiate between the two by looking at the
very first character: If it’s a slash, you’ve got an absolute filename (because the filename is
rooted to the very top level of the file system). If you don’t have a slash as the first character,
the filename’s a relative filename.
Earlier I told you that in the home directory at the top level of UNIX, I have a home directory
called taylor. In absolute filename terms, I’d properly say that I have /home/taylor as a
unique directory.
JUST A MINUTE
To add to the confusion, most UNIX people don’t pronounce the slashes,
particularly if the first component of the filename is a well-known directory.
I would pronounce /home/taylor as “home taylor,” but I would usually
pronounce /newt/awk/test as “slash newt awk test.” When in doubt,
pronounce the slash.
As you learn more about UNIX, particularly about how to navigate in the file system, you’ll
find that a clear understanding of the difference between a relative and absolute filename
proves invaluable. The rule of thumb is that if a filename begins with /, it’s absolute.
Task 3.2: Hidden Files in UNIX
One of the best aspects of living in an area for a long time, frequenting the same shops
and visiting the same restaurants, is that the people who work at each place learn your
name and preferences. Many UNIX applications can perform the same trick, remembering
your preferred style of interaction, what files you last worked with, which lines you’ve edited,
and more, through preference files.
On the Macintosh, because it’s a single-user system, there’s a folder within the System Folder
called Preferences, which is a central storage place for preference files, organized by
application. On my Macintosh, for example, I have about 30 different preference files in this
directory, enabling me to teach programs one time the defaults I prefer.
3
53
UNIX needs to support many users at once, so UNIX preference files can’t be stored in a
central spot in the file system. Otherwise, how would the system distinguish between your
preferences and those of your colleagues? To avoid this problem, all UNIX applications store
their preference files in your home directory.
Programs want to be able to keep their own internal preferences and status stored in your
directory, but these aren’t for you to work with or alter. If you use DOS, you’re probably
familiar with how the DOS operating system solves this problem: Certain files are hidden and
do not show up when you use DIR to list files in a directory.
Macintosh people don’t realize it, but the Macintosh also has lots of hidden files. On the
topmost level of the Macintosh file system, for example, the following files are present, albeit
hidden from normal display: AppleShare PDS, Deleted File Record, Desktop, Desktop DB,
and Desktop DF. Displaying hidden files on the Macintosh is very difficult, as it is with DOS.
Fortunately, the UNIX rule for hiding files is much easier than that for either the Mac or PC.
No secret status flag reminds the system not to display the file when listing directories.
Instead, the rule is simple, any filename starting with a dot. These files are called dot files.
A hidden file is any file with a dot as the first character of the filename.
JUST A MINUTE
If the filename or directory name begins with a dot, it won’t show up in normal listings of
that directory. If the filename or directory name has any other character as the first character
of the name, it lists normally.
1. Knowing that, turn to your computer and enter the ls command to list all the files
and directories in your home directory.
% ls –C -F
Archives/
InfoWorld/
LISTS
%
Mail/
News/
OWL/
RUMORS.18Sept
bin/
iecc.list
mailing.lists
newlists
src/
2. You can see that I have 12 items in my own directory, seven directories (the
directory names have a slash as the last character, remember) and five files. Files
have minimal rules for naming, too. Avoid slashes, spaces, and tabs, and you’ll
be fine.
3
Hour 3
54
3. Without an explicit direction to the contrary, UNIX is going to let the hidden files
remain hidden. To add the hidden files to the listing, you just need to add a -a flag
to the command. Turn to your computer and try this command to see what
hidden files are present in your directory. These are my results:
% ls -a
./
../
.Agenda
.aconfigrc
.article
.cshrc
.elm/
%
.gopherrc
.history*
.info
.letter
.login
.mailrc
.newsrc
.oldnewsrc
.plan
.pnewsexpert
.report
.rm-timestamp
.rnlast
.rnsoft
.sig
Archives/
InfoWorld/
LISTS
Mail/
News/
OWL/
RUMORS.18Sep
bin/
iecc.list
mail.lists
newlists
src/
Many dot files tend to follow the format of a dot, followed by the name of the
program that owns the file, with rc as the suffix. In my directory, you can see six
dot files that follow this convention: .aconfigrc, .cshrc, .gopherrc, .mailrc,
.newsrc, and .oldnewsrc .
Because of the particular rules of hidden files in UNIX, they are often called dot files,
and you can see that I have 23 dot files and directories in my directory.
JUST A MINUTE
JUST A MINUTE
The rc suffix tells you that this file is a configuration file for that particular
utility. For instance, .cshrc is the configuration file for the C shell and is
executed every time the C shell (/bin/csh) is executed. You can define
aliases for C shell commands and a special search path, for example.
Because it’s important to convey the specific filename of a dot file, pronunciation is a little different than elsewhere in UNIX. The name .gopherrc
would be spoken as “dot gopher are sea,” and .mailrc would be “dot mail
are sea.” If you can’t pronounce the program name, odds are good that no
one else can either, so .cshrc is “dot sea ess aitch are sea.”
Other programs create a bunch of different dot files and try to retain a consistent naming
scheme. You can see that .rnlast and .rnsoft are both from the rn program, but it’s difficult
to know simply from the filenames that .article, .letter, .newsrc, .oldnewsrc, and
.pnewsexpert are all also referenced by the rn program. Recognizing this problem, some
application authors designed their applications to create a dot directory, with all preference
files neatly tucked into that one spot. The elm program does that with its .elm hidden
directory.
3
55
Some files are directly named after the programs that use them: the .Agenda file is used by the
agenda program, and .info is used by the info program. Those almost have a rule of their
own, but it’s impossible to distinguish them from .login, from the sh program; .plan from
the finger program; .rm-timestamp from a custom program of my own; and I frankly have
no idea what program created the .report file!
This should give you an idea of the various ways that UNIX programs name and use hidden
files. As an exercise, list all the dot files in your home directory and try to extract the name
of the program that probably created the file. Check by looking in the index of this book to
see if a program by that name exists. If you can’t figure out which programs created which
files, you’re not alone. Keep the list handy; refer to it as you learn more about UNIX while
exploring Teach Yourself UNIX in 24 Hours, and by the time you’re done, you’ll know exactly
how to find out which programs created which dot files.
Task 3.3: The Special Directories “.” and “..”
There are two dot directories I haven’t mentioned, although they show up in my
listing and most certainly show up in your listing, too. They are dot and dot dot (“.”
and “..”), and they’re shorthand directory names that can be terrifically convenient.
The dot directory is shorthand for the current location in the directory hierarchy; the dot-dot
directory moves you up one level, to the parent directory.
Consider again the list of files shown in Figure 3.3. If you were looking at the files in the
California Personnel directory (best specified as /California/Personnel) and wanted to
check quickly an entry in the Bldgs file for California, either you’d have to use the absolute
filename and enter the lengthy ls /California/Bldgs, or, with the new shorthand directories,
you could enter ls ../Bldgs.
As directories move ever deeper into the directory hierarchy, the dot-dot notation can save
you much typing time. For example, what if the different states and related files were all
located in my home directory /home/taylor, in a new directory called business? In that case,
the absolute filename for employee Raab,M. in California would be /home/taylor/business/
California/Personnel/Raab,M. , which is unwieldy and an awful lot to type if you want to
hop up one level and check on the buildings database in Indiana!
You can use more than one dot-dot notation in a filename, too, so if you’re looking at the
Raab,M. file and want to check on Dunlap,L. , you could save typing in the full filename by
instead using ../../../Washington/Personnel/Dunlap,L.. Look at Figure 3.3 to see how
that would work, tracing back one level for each dot-dot in the filename.
This explains why the dot-dot shorthand is helpful, but what about the single-dot notation
that simply specifies the current directory?
3
56
Hour 3
I haven’t stated it explicitly yet, but you’ve probably figured out that one ramification of the
UNIX file system organization, combined with its capability to place applications anywhere
in the file system, is that the system needs some way to know where to look for particular
applications. Just as if you were looking for something in a public library, in UNIX, having
an understanding of its organization and a strategy for searching is imperative for success
and speed.
UNIX uses an ordered list of directories called a search path for this purpose. The search path
typically lists five or six different directories on the system where the computer checks for any
application you request.
The question arises: What happens if your own personal copy of an application has the same
name as a standard system application? The answer is that the system always finds the
standard application first, if its directory is listed earlier in the search path.
To avoid this pitfall, you need to use the dot notation, forcing the system to look in the current
directory rather than search for the application. If you wanted your own version of the ls
command, for example, you’d need to enter ./ls to ensure that UNIX uses your version
rather than the standard version.
1. Enter ./ls on your computer and watch what happens.
2. Enter ls without the dot notation, and notice how the computer searches through
various directories in the search path, finds the ls program, and executes it,
automatically.
When you learn about cd later in the book, you also will learn other uses of the dotdot directory, but the greatest value of the dot directory is that you can use it to force
the system to look in the current directory and nowhere else for any file specified.
Task 3.4: The env Command
You’ve learned a lot of the foundations of the UNIX file system and how applications
remember your preferences through hidden dot files. There’s another way, however,
that the system remembers specifics about you, and that’s through your user environment.
The user environment is a collection of specially named variables that have specific values.
3
57
1. To view your environment, you can use the env command. Here’s what I see when
I enter the env command on my system:
% env
HOME=/users/taylor
SHELL=/bin/csh
TERM=vt100
PATH=/users/taylor/bin:/bin:/usr/bin:/usr/ucb:/usr/local/bin:
➥/usr/unsup/bin:.
MAIL=/usr/spool/mail/taylor
LOGNAME=taylor
TZ=EST5
%
Try it yourself and compare your values with mine. You might find that you have
more defined in your environment than I do because your UNIX system uses your
environment to keep track of more information.
JUST A MINUTE
Many UNIX systems offer the printenv command instead of env. If you
enter env and the system complains that it can’t find the env command, try
using printenv instead. All examples here work with either env or
printenv.
Task 3.5: PATH and HOME
The two most important values in your environment are the name of your home
directory (HOME) and your search path (PATH). Your home directory (as it’s known)
is the name of the directory that you always begin your UNIX session within.
The PATH environment variable lists the set of directories, in left-to-right order, that the
system searches to find commands and applications you request. You can see from the
example that my search path tells the computer to start looking in the /users/taylor/bin
directory, then sequentially try /bin, /usr/bin, /usr/ucb, /usr/local/bin, /usr/unsup/bin,
and . before concluding that it can’t find the requested command. Without a PATH, the shell
wouldn’t be able to find any of the many, many UNIX commands: As a minimum, you
always should have /bin and /usr/bin.
3
Hour 3
58
1. You can use the echo command to list specific environment variables, too. Enter
echo $PATH and echo $HOME. When I do so, I get the following results:
% echo $PATH
/users/taylor/bin:/bin:/usr/bin:/usr/ucb:/usr/local/bin:/usr/unsup/bin:.
% echo $HOME
/users/taylor
%
Your PATH value is probably similar, although certainly not identical, to mine, and
your HOME is /home/accountname or similar (accountname is your account name).
Task 3.6: Find Where You Are with pwd
So far you’ve learned a lot about how the file system works but not much about how
to move around in the file system. With any trip, the first and most important step
is to find out your current location—that is the directory in which you are currently working.
In UNIX, the command pwd tells you the present working directory.
1. Enter pwd. The output should be identical to the output you saw when you entered
env HOME because you’re still in your home directory.
% env HOME
/users/taylor
% pwd
/users/taylor
%
Think of pwd as a compass, always capable of telling you where you are. It also tells
you the names of all directories above you because it always lists your current location
as an absolute directory name.
Task 3.7: Move to Another Location with cd
The other half of the dynamic duo is the cd command, which is used to change
directories. The format of this command is simple, too: cd new-directory (where
new-directory is the name of the new directory you want).
3
59
1. Try moving to the very top level of the file system and entering pwd to see if the
computer agrees that you’ve moved.
% cd /
% pwd
/
%
2. Notice that cd doesn’t produce any output. Many UNIX commands operate
silently like this, unless an error is encountered. The system then indicates the
problem. You can see what an error looks like by trying to change your location to
a nonexistent directory. Try the /taylor directory to see what happens!
% cd /taylor
/taylor: No such file or directory
%
3. Enter cd without specifying a directory. What happens? I get the following result:
% cd
% pwd
/users/taylor
%
4. Here’s where the HOME environment variable comes into play. Without any directory specified, cd moves you back to your home directory automatically. If you get
lost, it’s a fast shorthand way to move to a known location without fuss.
Remember the dot-dot notation for moving up a level in the directory hierarchy?
Here’s where it also proves exceptionally useful. Use the cd command without any
arguments to move to your home directory, then use pwd to ensure that’s where
you’ve ended up.
5. Now, move up one level by using cd .. and check the results with pwd:
% cd
% pwd
/users/taylor
% cd ..
% pwd
/users
%
3
Hour 3
60
6. Use the ls -C -F command to list all the directories contained at this point in the
file system. Beware, though; on large systems, this directory could easily have
hundreds of different directories. On one system I use, there are almost 550
different directories one level above my home directory in the file system!
% ls -C -F
armstrong/
bruce/
cedric/
%
christine/
david/
green/
guest/
laura/
higgins/ mac/
kane/
mark/
matthewm/ shane/
rank/
taylor/
shalini/ vicki/
Try using a combination of cd and pwd to move about your file system, and remember
that without any arguments, cd always zips you right back to your home directory.
Summary
This hour has focused on the UNIX hierarchical file system. You’ve learned the organization
of a hierarchical file system, how UNIX differs from Macintosh and DOS systems, and how
UNIX remembers preferences with its hidden dot files. This hour has also explained the
difference between relative and absolute filenames, and you’ve learned about the “.” and “..”
directories. You’ve learned three new commands too: env to list your current environment,
cd to change directories, and pwd to find out your present working directory location.
Workshop
in the next hour.
Key Terms
absolute filename Any filename that begins with a leading slash (/); these always uniquely
describe a single file in the file system.
binary A file format that is intended for the computer to work with directly rather than for
humans to peruse. See also executable.
device driver All peripherals attached to the computer are called devices in UNIX, and
each has a control program always associated with it, called a device driver. Examples are the
device drivers for the display, keyboard, mouse, and all hard disks.
directory A type of UNIX file used to group other files. Files and directories can be placed
inside other directories, to build a hierarchical system.
3
61
directory separator character On a hierarchical file system, there must be some way to
specify which items are directories and which is the actual filename itself. This becomes
particularly true when you’re working with absolute filenames. In UNIX, the directory
separator character is the slash (/), so a filename like /tmp/testme is easily interpreted as a file
called testme in a directory called tmp.
dot A shorthand notation for the current directory.
dot-dot A shorthand notation for the directory one level higher up in the hierarchical file
system from the current location.
dot file A configuration file used by one or more programs. These files are called dot files
because the first letter of the filename is a dot, as in .profile or .login. Because they’re dot
files, the ls command doesn’t list them by default, making them also hidden files in UNIX.
See also hidden file.
dynamic linking Although most UNIX systems require all necessary utilities and library
routines (such as the routines for reading information from the keyboard and displaying it
to the screen) to be plugged into a program when it’s built (known in UNIX parlance as static
linking), some of the more sophisticated systems can delay this inclusion until you actually
need to run the program. In this case, the utilities and libraries are linked when you start the
program, and this is called dynamic linking.
executable A file that has been set up so that UNIX can run it as a program. This is also
shorthand for a binary file. You also sometimes see the phrase binary executable, which is the
same thing! See also binary.
hidden file By default, the UNIX file-listing command ls shows only files whose first letter
isn’t a dot (that is, those files that aren’t dot files). All dot files, therefore, are hidden files, and
you can safely ignore them without any problems. Later, you learn how to view these hidden
files. See also dot file.
home directory This is your private directory, and is also where you start out when you
log in to the system.
kernel The underlying core of the UNIX operating system itself. This is akin to the
concrete foundation under a modern skyscraper.
preference file These are what dot files (hidden files) really are: They contain your
individual preferences for many of the UNIX commands you use.
relative filename Any filename that does not begin with a slash (/) is a filename whose exact
meaning depends on where you are in the file system. For example, the file test might exist
in both your home directory and in the root directory; /test is an absolute filename and leaves
no question which version is being used, but test could refer to either copy, depending on
your current directory.
3
Hour 3
62
root directory
The directory at the very top of the file system hierarchy, also known as slash.
search path A list of directories used to find a command. When a user enters a command
ls, the shell looks in each directory in the search path to find a file ls, either until it is found
or the list is exhausted.
slash The root directory.
symbolic link A file that contains a pointer to another file rather than contents of its own.
This can also be a directory that points to another directory rather than having files of its own.
A useful way to have multiple names for a single program or allow multiple people to share
a single copy of a file.
user environment A set of values that describe the user’s current location and modify the
behavior of commands.
working directory The directory where the user is working.
Questions
1. Can you think of information you work with daily that’s organized in a hierarchical
fashion? Is a public library organized hierarchically?
2. Which of the following files are hidden files and directories according to UNIX?
.test
../
hide-me
.dot.
,test
dot
.cshrc
.HiMom
3. What programs most likely created the following dot files and dot directories?
.cshrc
.tmp334
.rnsoft
.excel/
.exrc
.letter
.print
.vi-expert
4. In the following list, circle the items that are absolute filenames:
/Personnel/Taylor,D.
/home/taylor/business/California
../..
Recipe:Gazpacho
5. Using the list of directories found on all UNIX systems (/bin, /dev, /etc, /lib,
/lost+found , /mnt, /sys, /tmp, /usr ), use cd and pwd to double-check that they are
all present on your own UNIX machine.
In the next hour, you learn about the ls command that you’ve been using, including a further
discussion of command flags. The command touch enables you to create your own files, and
du and df help you learn how much disk space is used and how much is available, respectively.
You also learn how to use two valuable if somewhat esoteric UNIX commands, compress and
crypt, which help you minimize your disk-space usage and ensure absolute security for
special files.
3
Hour
4
Listing Files and
Managing Disk Usage
This hour introduces you to the ls command, one of the most commonly used
commands in UNIX. The discussion includes over a dozen different command
options, or flags. You also learn how to use the touch command to create files,
how to use the du command to see how much disk space you’re using, and how
to use the df command to see how much disk space is available. Finally, the
compress command can help you minimize your disk-space usage, particularly
on files you’re not using very often.
Goals for This Hour
■
■
■
■
■
■
63
All about the ls command
About special ls command flags
How to create files with touch
How to check disk-space usage with du
How to check available disk space with df
How to shrink big files with compress
4
Hour 4
64
Your first hours focused on some of the basic UNIX commands, particularly those for
interacting with the system to accomplish common tasks. In this hour, you expand that
knowledge by analyzing characteristics of the system you’re using, and you learn a raft of
commands that let you create your own UNIX workspace. You also learn more about the
UNIX file system and how UNIX interprets command lines. In addition to the cd and pwd
commands that you learned in the preceding hour, you learn how to use ls to wander in the
file system and see what files are kept where.
Unlike the DOS and Macintosh operating systems, information about the UNIX system is
often difficult to obtain. In this hour, you learn easy ways to ascertain how much disk space
you’re using, with the du command. You also learn how to interpret the oft-confusing output
of the df command, which enables you to see instantly how much total disk space is available
on your UNIX system.
This hour concludes with a discussion of the
shrink the size of any file or set of files.
compress
command, which enables you to
The ls Command
This section introduces you to the ls command, which enables you to wander in the file
system and see what files are kept where.
Task 4.1: All About the ls Command
From the examples in the previous hour, you’ve already figured out that the
command used to list files and directories in UNIX is the ls command.
All operating systems have a similar command, a way to see what’s in the current location.
In DOS, for example, you’re no doubt familiar with the DIR command. DOS also has
command flags, which are denoted by a leading slash before the specific option. For example,
DIR /W produces a directory listing in wide-display format. The DIR command has quite a few
other options and capabilities.
Listing the files in a directory is a pretty simple task, so why all the different options? You’ve
already seen some examples, including ls -a, which lists hidden dot files. The answer is that
there are many different ways to look at files and directories, as you will learn.
1. The best way to learn what ls can do is to go ahead and use it. Turn to your
computer, log in to your account, and try each command as it’s explained.
4
65
2. The most basic use of ls is to list files. The command ls lists all the files and
directories in the present working directory (recall that you can check what
directory you’re in with the pwd command at any time).
% ls
Archives
InfoWorld
LISTS
Mail
News
OWL
RUMORS.18Sept
bin
iecc.list
mailing.lists
newels
src
Notice that the files are sorted alphabetically from top to bottom, left to right. This
is the default, known as column-first order because it sorts downward, then across.
You should also note how things are sorted in UNIX: The system differentiates
between uppercase and lowercase letters, unlike DOS. (The Macintosh remembers
whether you use uppercase or lowercase letters for naming files, but it can’t
distinguish between them internally. Try it. Name one file TEST and another file
test the next time you’re using a Macintosh.)
JUST A MINUTE
Some of the UNIX versions available for the PC—notably SCO and
INTERACTIVE UNIX—have an ls that behaves slightly differently and may
list all files in a single column rather than in multiple columns. If your PC
does this, you can use the -C flag to ls to force multiple columns.
It’s important that you always remember to type UNIX commands in lowercase
letters, unless you know that the particular command is actually uppercase; remember
that UNIX treats Archives and archives as different filenames. Also, avoid entering your
account name in uppercase when you log in. UNIX has some old compatibility features that
make using the system much more difficult if you use an all-uppercase login. If you ever
accidentally log in with all uppercase, log out and try again in lowercase.
Task 4.2: Having ls Tell You More
Without options, the ls command offers relatively little information. Questions you
might still have about your directory include: How big are the files? Which are files,
and which are directories? How old are they? What hidden files do you have?
1. Start by entering ls
% ls -s
total 403
1 Archives
1 InfoWorld
108 LISTS
-s
to indicate file sizes:
1 Mail
1 News
1 OWL
5 RUMORS.18Sept
1 bin
4 iecc.list
280 mailing.lists
2 newels
1 src
4
Hour 4
66
2. To ascertain the size of each file or directory listed, you can use the -s flag to ls.
The size indicated is the number of kilobytes, rounded upward, for each file. The
first line of the listing also indicates the total amount of disk space used, in kilobytes, for the contents of this directory. The summary number does not, however,
include the contents of any subdirectories, so it’s deceptively small.
JUST A MINUTE
A kilobyte is 1,024 bytes of information, a byte being a single character.
The preceding paragraph, for example, contains slightly more than 400
characters. UNIX works in units of a block of information, which, depending on which version of UNIX you’re using, is either 1 kilobyte or 512
bytes. Most UNIX systems now work with a 1-kilobyte block.
3. Here is a further definition of what occurs when you use the -s flag: ls indicates
the number of blocks each file or directory occupies. You then can use simple
calculations to convert blocks into bytes. For example, the ls command indicates
that the LISTS file in my home directory occupies 108 blocks. A quick calculation
of block size multiplied by the number of blocks reveals the actual file size, in bytes,
of LISTS, as shown here:
% bc
1024 * 108
110592
quit
%
Based on these results of the bc command, you can see that the file is 110,592 bytes
in size. You can estimate size by multiplying the number of blocks by 1,000. Be
aware, however, that in large files, the difference between 1,000 and 1,024 is
significant enough to introduce an error into your calculation. As an example, I
have a file that’s more than three megabytes in size (a megabyte is 1,024 kilobytes,
which is 1,024 bytes, so a megabyte is 1,024 × 1,024, or 1,048,576 bytes):
% ls -s bigfile
3648 bigfile
4. The file actually occupies 3,727,360 bytes. If I estimated its size by multiplying the
number of blocks by 1,000 (which equals 3,648,000 bytes), I’d have underestimated its size by 79,360 bytes. (Remember, blocks × 1,000 is an easy estimate!)
JUST A MINUTE
4
The last example reveals something else about the ls command. You can
specify individual files or directories you’re interested in viewing and
avoid having to see all files and directories in your current location.
67
5. You can specify as many files or directories as you like, and separate them by
spaces:
% ls -s LISTS iecc.list newels
108 LISTS
4 iecc.list
2 newels
In the previous hour, you learned that UNIX identifies each file that begins with a
dot (.) as a hidden file. Your home directory is probably littered with dot files,
which retain preferences, status information, and other data. To list these hidden
files, use the -a flag to ls:
% ls -a
.
..
.Agenda
.aconfigrc
.article
.cshrc
.elm
.gopherrc
.history
.info
.letter
.login
.mailrc
.newsrc
.oldnewsrc
.plan
.pnewsexpert
.report
.rm-timestamp
.rnlast
.rnsoft
.sig
Archives
InfoWorld
LISTS
Mail
News
OWL
RUMORS.18Sept
bin
iecc.list
mailing.lists
newels
src
You can see that this directory contains more dot files than regular files and
directories. That’s not uncommon in a UNIX home directory. However, it’s rare to
find any dot files other than the standard dot and dot-dot directories (those are in
every directory in the entire file system) in directories other than your home
directory.
6. You used another flag to the ls command—the -F flag—in the previous hour. Do
you remember what it does?
% ls -F
Archives/
InfoWorld@
LISTS
Mail/
News/
OWL/
RUMORS.18Sept
bin/
iecc.list
mailing.lists
newels
src/
Adding the -F flag to ls appends suffixes to certain filenames so that you can
ascertain more easily what types of files they are. Three different suffixes can be
added, as shown in Table 4.1.
Table 4.1. Filename suffixes appended by ls -F.
Suffix
Example
Meaning
/
Mail/
Mail
*
prog*
@
bin@
is a directory.
prog is an executable program.
bin is a symbolic link to another file or directory.
7. If you’re familiar with the Macintosh and have used either System 7.0 or 7.1, you
may recall the new feature that enables the user to create and use an alias. An alias
is a file that does not contain information, but acts, instead, as a pointer to the
actual information files. Aliases can exist either for specific files or for folders.
4
Hour 4
68
UNIX has offered a similar feature for many years, which in UNIX jargon is called
a symbolic link. A symbolic link, such as bin in Table 4.1, contains the name of
another file or directory rather than any contents of its own. If you could peek
inside, it might look like bin = @/usr/bin. Every time someone tries to look at bin,
the system shows the contents of /usr/bin instead.
You’ll learn more about symbolic links and how they help you organize your files a
bit later in the book. For now, just remember that if you see an @ after a filename,
it’s a link to another spot in the file system.
8. A useful flag for ls (one that might not be available in your version of UNIX) is
the -m flag. This flag outputs the files as a comma-separated list. If there are many
files, -m can be a quick and easy way to see what’s available.
% ls -m
Archives, InfoWorld, LISTS, Mail, News, OWL, RUMORS.18Sept,
bin, iecc.list, mailing.lists, newels, src
Sometime you might want to list each of your files on a separate line, perhaps for a
printout you want to annotate. You’ve seen that the -C flag forces recalcitrant versions
of ls to output in multiple columns. Unfortunately, the opposite behavior isn’t obtained
using a lowercase c. (UNIX should be so consistent!) Instead, use the -1 flag to indicate that
you want one column of output. Try it.
Task 4.3: Combining Flags
The different flags you’ve learned so far are summarized in Table 4.2.
Table 4.2. Some useful flags to ls.
Flag
Meaning
-a
List all files, including any dot files.
Indicate file types; / = directory, * = executable.
Show files as a comma-separated list.
Show size of files, in blocks (typically, 1 block = 1,024 bytes).
Force multiple-column output on listings.
Force single-column output on listings.
-F
-m
-s
-C
-1
What if you want a list, generated with the -F conventions, that simultaneously shows you
all files and indicates their types?
4
69
1. Combining flags in UNIX is easy. All you have to do is run them together in a
sequence of characters, and prefix the whole thing with a dash:
% ls -aF
./
.gopherrc
➥RUMORS.18Sept
../
.history*
➥bin/
.Agenda
.info
➥iecc.list
.aconfigrc
.letter
➥mailing.lists
.article
.login
➥newels
.cshrc
.mailrc
➥src/
.elm/
.newsrc
.oldnewsrc
.sig
.plan
Archives/
.pnewsexpert
InfoWorld/
.report
LISTS
.rm-timestamp
Mail/
.rnlast
News/
.rnsoft
OWL/
2. Sometimes it’s more convenient to keep all the flags separate. This is fine, as long as
each flag is prefixed by its own dash:
% ls -s -F
total 403
1 Archives/
1 InfoWorld/
108 LISTS
1 Mail/
1 News/
1 OWL/
5 RUMORS.18Sept
1 bin/
4 iecc.list
280 mailing.lists
2 newels
1 src/
3. Try some of these combinations on your own computer. Also try to list a flag more
than once (for example, ls -sss -s), or list flags in different orders.
Very few UNIX commands care about the order in which flags are listed. Because it’s
the presence or absence of a flag that’s important, listing a flag more than once doesn’t
make any difference.
Task 4.4: Listing Directories Without Changing
Location
Every time I try to do any research in the library, I find myself spending hours and
hours there, but it seems to me that I do less research than I think I should. That’s
because most of my time is for the tasks between the specifics of my research: finding the
location of the next book, and finding the book itself.
If ls constrained you to listing only the directory that you were in, it would hobble you in
a similar way. Using only ls would slow you down dramatically and force you to use cd to
move around each time.
Instead, just as you can specify certain files by using ls, you can specify certain directories
you’re interested in viewing.
4
Hour 4
70
1. Try this yourself. List /usr on your system:
% ls -F /usr
5bin/
5include/
5lib/
acc/
acctlog*
adm@
bin/
boot@
demo/
diag/
dict/
etc/
export/
games/
hack/
hosts/
include/
kvm/
lddrv/
lib/
local/
lost+found/
man@
mdec@
old/
pub@
sccs/
share/
source/
spool@
src@
stand@
sys@
system/
tmp@
ucb/
ucbinclude@
ucblib@
xpg2bin/
xpg2include/
xpg2lib/
You probably have different files and directories listed in your own /usr directory.
Remember, @ files are symbolic links in the listing, too.
2. You can also specify more than one directory:
% ls /usr/local /home/taylor
/home/taylor:
Global.Software
Mail/
Src/
Interactive.Unix News/
bin/
/usr/local/:
T/
emacs/
ftp/
admin/
emacs-18.59/ gnubin/
bin/
etc/
include/
cat/
faq/
info/
doc/
forms/
lib/
history.usenet.Z
lists/
lost+found/
man/
menu/
motd
motd~
netcom/
policy/
src/
tmp/
In this example, the ls command also sorted the directories before listing them. I
specified that I wanted to see /usr/local and then /home/taylor, but it presented
the directories in opposite order.
JUST A MINUTE
I’ve never been able to figure out how ls sorts directories when you ask
for more than one to be listed—it’s not an alphabetical listing. Consider it
a mystery. Remember that if you must have the output in a specific order,
you can use the ls command twice in a row.
3. Here’s where the dot-dot shorthand can come in handy. Try it yourself:
% ls -m ..
armstrong, bruce, cedric, christine, david, green,
guest, higgins, james, kane, laura, mac, mark,
patrickb, rank, shalini, shane, taylor, vicki
If you were down one branch of the file system and wanted to look at some files
down another branch, you could easily find yourself using the command ls ../
Indiana/Personnel or ls -s ../../source .
4
71
4. There’s a problem here, however. You’ve seen that you can specify filenames to
look at those files, and directory names to look at the contents of those directories,
but what if you’re interested in the directory itself, not in its contents? I might
want to list just two directories—not the contents, just the files themselves, as
shown here:
% ls -F
Archives/
InfoWorld/
LISTS
% ls -s LISTS
108 LISTS
Mail:
total 705
8 cennamo
28 dan_sommer
14 decc
3 druby
Mail/
News/
OWL/
Mail newlists
2 newlists
RUMORS.18Sept
bin/
iecc.list
27
2
48
14
4 kcs
21
34 lehman 5
64 mac
7
92 mailbox 5
ean_houts
gordon_haight
harrism
james
mailing.lists
newlists
src/
mark
7 sartin
raf
3 shelf
rock
20 steve
rustle 18 tai
5. The problem is that ls doesn’t know that you want to look at Mail unless you tell
it not to look inside the directories specified. The command flag needed is -d,
which forces ls to list directories rather than their contents. The same ls command, but with the -d flag, has dramatically different output:
% ls -ds LISTS Mail newlists
108 LISTS
1 Mail/
2 newlists
Try some of these flags on your own system, and watch how they work together.
To list a file or directory, you can specify it to ls. Directories, however, reveal their
contents, unless you also include the -d flag.
Special ls Command Flags
It should be becoming clear to you that UNIX is the ultimate toolbox. Even some of the
simplest commands have dozens of different options. On one system I use, ls has more than
20 different flags.
Task 4.5: Changing the Sort Order in ls
What if you wanted to look at files, but wanted them to show up in a directory sorting
order different from the default (that is, column-first order)? How could you change
the sort order in ls?
1. The -x flag sorts across, listing the output in columns, or first-row order (entries
are sorted across, then down):
4
Hour 4
72
% ls -a
.
..
.Pnews.header
.accinfo
.article
.cshrc
.delgroups
% ls -x -a
.
.article
.forward
.newsrc
.rnlast
.tin
News
.elm
.forward
.ircmotd
.login
.logout
.newsrc
.oldnewsrc
.plan
.pnewsexpert
.rnlast
.rnlock
.rnsoft
.sig
.tin
Global.Software
Interactive.Unix
Mail
News
Src
bin
history.usenet.Z
..
.cshrc
.ircmotd
.oldnewsrc
.rnlock
Global.Software
Src
.Pnews.header
.delgroups
.login
.plan
.rnsoft
Interactive.Unix
bin
.accinfo
.elm
.logout
.pnewsexpert
.sig
Mail
history.usenet.Z
2. There are even more ways to sort files in ls. If you want to sort by most-recentlyaccessed to least-recently-accessed, you use the -t flag:
% ls -a -t
./
.newsrc
.oldnewsrc
.article
.elm/
.forward
history.usenet.Z
../
News/
.tin/
.ircmotd
.delgroups
.login
bin/
.rnlock
.rnlast
.rnsoft
Interactive.Unix
.accinfo*
Src/
Global.Software
.cshrc
.sig
.plan
Mail/
.Pnews.header*
.pnewsexpert
.logout
From this output, you can see that the most recently accessed files are .newsrc and
.oldnewsrc , and that it’s been quite a while since .logout was touched. Try using
the -t flag on your system to see which files you’ve been accessing and which you
haven’t.
3. So far, you know three different approaches to sorting files within the ls command: column-first order, row-first order, and most-recently-accessed-first order.
But there are more options in ls than just these three; the -r flag reverses any
sorting order.
% ls
Global.Software
Interactive.Unix
% ls -r
history.usenet.Z
bin/
Mail/
News/
Src/
bin/
history.usenet.Z
Src/
News/
Mail/
Interactive.Unix
Global.Software
4. Things may become confusing when you combine some of these flags. Try to list
the contents of the directory that is one level above the current directory, sorted so
the most-recently-accessed file is last in the list. At the same time, indicate which
items are directories and the size of each file.
% ls -r -t -F -s ..
total 150
2 bruce/
2 rank/
4
2 kane/
14 higgins/
2
2
2
2
laura/
cedric
james@
vicki/
2
2
4
2
christine/
peggy/
taylor/
guest/
2
4
4
6
shane/
patrickb/
green/
shalini/
73
6 mac/
10 mark/
6 armstrong/
4 david/
A better, easier way to type the previous command would be to bundle flags into the
single argument ls -rtFs .., which would work just as well, and you’d look like an
expert!
Task 4.6: Listing Directory Trees Recursively in ls
In case things aren’t yet complicated enough with ls, two more important, valuable
flags are available. One is the -R flag, which causes ls to list recursively directories
below the current or specified directory. If you think of listing files as a numbered set of steps,
recursion is simply adding a step—the rule is if this file is a directory, list it, too—to the list.
1. When I use the -R flag, here’s what I see:
% ls -R
Global.Software
Mail/
Src/
history.usenet.Z
Interactive.Unix
News/
bin/
Mail:
Folders/ Netnews/
Mail/Folders:
mail.sent mailbox
steinman
tucker
Mail/Netnews:
postings
News:
uptodate volts
Src:
sum-up.c
bin:
Pnews*
punt*
submit*
Try it yourself.
Notice that ls lists the current directory and then alphabetically lists the contents
of all subdirectories. Notice also that the Mail directory has two directories within
it and that those are also listed here.
Viewing all files and directories below a certain point in the file system can be a
valuable way to look for files (although you’ll soon learn better tools for finding files).
If you aren’t careful, though, you may get hundreds or thousands of lines of information
streaming across your screen. Do not enter a command like ls -R / unless you have time to
sit and watch information fly past.
4
Hour 4
74
If you try to list the contents of a directory when you don’t have permission to access the
information, ls warns you with an error message:
% ls ../marv
../marv unreadable
Now ask for a recursive listing, with indications of file type and size, of the directory /etc,
and see what’s there. The listing will include many files and subdirectories, but they should
be easy to wade through due to all the notations ls uses to indicate files and directories.
Task 4.7: Long Listing Format in ls
You’ve seen how to estimate the size of a file by using the -s flag to find the number
of blocks it occupies. To find the exact size of a file in bytes, you need to use the
-l flag. (Use a lowercase letter L. The numeral 1 produces single-column output, as you’ve
already learned.)
1. The first long listing shows information for the LISTS file.
% ls -l LISTS
-rw------- 1 taylor
106020 Oct
8 15:17 LISTS
The output is explained in Figure 4.1.
Figure 4.1.
The meaning of the -l
output for a file.
For each file and directory in the UNIX file system, the owner, size, name, number
of other files pointing to it (links), and access permissions are recorded. The creation,
modification, and access times and dates are also recorded for each file. The modification
time is the default time used for the -t sorting option and listed by the ls long format.
Permissions Strings
Interpreting permissions strings is a complex issue because UNIX has a sophisticated security
model for individual files. Security revolves around three different types of users: the owner
of the file, the group of which that the file is a part, and everyone else.
4
75
The first character of the permissions string, identified in Figure 4.1 as access permissions,
indicates the kind of file. The two most common values are d for directories and - for regular
files. Be aware that there are many other file types that you’ll rarely, if ever, see.
The following nine characters in the permissions string indicate what type of access is allowed
for different users. From left to right, these characters show what access is allowed for the
owner of the file, the group that owns the file, and everyone else.
Figure 4.2 shows how to break down the permissions string for the LISTS file into individual
components.
Figure 4.2.
Reading access permissions for LISTS .
Each permissions string is identically composed of three components—permission for
reading, writing, and execution—as shown in Figure 4.3.
Figure 4.3.
Elements of a permissions
string.
Armed with this information—specifically, knowing that a - character means that the
specific permission is denied—you can see that ls shows that the owner of the file, taylor,
as illustrated in Figure 4.1, has read and write permission. Nobody else either in taylor’s
group or in any other group has permission to view, edit, or run the file.
Earlier you learned that just about everything in UNIX ends up as a file in the file system,
whether it’s an application, a device driver, or a directory. The system keeps track of whether
a file is executable because that’s one way it knows whether LISTS is the name of a file or the
name of an application.
Task 4.8: Long Listing Format for Directories in ls
The long form of a directory listing is almost identical to a file listing, but the
permissions string is interpreted in a very different manner.
4
Hour 4
76
1. Here is an example of a long directory listing:
% ls -l -d Example
drwxr-x--- 2 taylor
1024 Sep 30 10:50 Example/
Remember that you must have both read and execute permission for a directory. If
you have either read or execute permission but not both, the directory will not be
usable (as though you had neither permission). Write permission, of course, enables
the user to alter the contents of the directory or add new files to the directory.
2. The Example directory breaks down for interpretation as shown in Figure 4.4.
Figure 4.4.
Elements of directory
permissions.
JUST A MINUTE
I’ve never understood the nuances of a directory with read but not execute
permission, or vice versa, and explanations from other people have never
proven to be correct. It’s okay, though, because I’ve never seen a
directory on a UNIX system that was anything other than ---, r-x, or rwx.
3. Now try using the -l flag yourself. Move to your home directory, and enter ls
as shown here:
% ls -l
total 403
drwx-----drwx------
4
2 taylor
3 taylor
512 Sep 30 10:38 Archives/
512 Oct 1 08:23 InfoWorld/
-l
-rw------drwx-----drwx-----drwx------rw------drwx------rw-------rw-rw----rw-rw---drwx------
1
2
2
2
1
2
1
1
1
2
taylor
taylor
taylor
taylor
taylor
taylor
taylor
taylor
taylor
taylor
106020
1024
512
512
4643
512
3843
280232
1031
512
Oct
Sep
Oct
Sep
Sep
Oct
Oct
Oct
Oct
Sep
8
30
6
30
20
1
6
6
7
14
15:17
10:50
09:36
10:51
10:49
09:53
18:02
09:57
15:44
22:14
77
LISTS
Mail/
News/
OWL/
RUMORS.18Sept
bin/
iecc.list
mailing.lists
newlists
src/
The size of a directory is usually in increments of 512 bytes. The second field, the
“link,” is an interesting and little-known value when a directory is being listed.
Instead of counting up the number of other files that point to the file, (that is, the
number of files that have a link to the current file), the second field indicates the
number of directories that are contained in that specific directory. Remember, all
directories have dot and dot-dot, so the minimum value is always 2.
4. Consider the following example:
% ls -Fa
./
.gopherrc
../
.history*
➥RUMORS.18Sept
.Agenda
.info
.aconfigrc
.letter
.article
.login
➥mailing.lists
.cshrc
.mailrc
.elm/
.newsrc
% ls -ld .
drwx------ 10 taylor
.oldnewsrc
.plan
.sig
Archives/
OWL/
.pnewsexpert
.report
.rm-timestamp
Cancelled.mail
InfoWorld/
LISTS
bin/
iecc.list
.rnlast
.rnsoft
Mail/
News/
newlists
src/
1024 Oct 10 16:00 ./
5. Try entering ls -ld. and see if it correctly identifies the number of directories in
your home directory. Move to other directories and see whether the listing agrees
with your own count of directories.
The output from the ls -l command is unquestionably complex and packed with
information. Interpretation of permissions strings is an important part of understanding and being able to use UNIX, and more explanation is offered in subsequent hours.
Table 4.3 summarizes the many different command flags for ls that you have learned in this
hour.
Table 4.3. Summary of command flags for ls.
Flag
Meaning
-1
Force single-column output on listings.
List all files, including any dot files.
Force multiple-column output on listings.
-a
-C
continues
4
Hour 4
78
Table 4.3. continued
Flag
Meaning
-d
List directories rather than their contents.
Indicate file types; / = directory, * = executable.
Generate a long listing of files and directories.
Show files as a comma-separated list.
Reverse the order of any file sorting.
Recursively show directories and their contents.
Show size of files, in blocks (typically 1 block = 1,024 bytes).
Sort output in most-recently-modified order.
Sort output in row-first order.
-F
-l
-m
-r
-R
-s
-t
-x
Without doubt, ls is one of the most powerful and, therefore, also one of the most confusing
commands in UNIX. The best way for you to learn how all the flags work together is to
experiment with different combinations.
Task 4.9: Creating Files with the touch Command
At this point, you have a variety of UNIX tools that help you move through the file
system and learn about specific files. The touch command is the first command that
helps you create new files on the system, independent of any program other than the shell
itself. This can prove very helpful for organizing a new collection of files, for example.
The main reason that touch is used in UNIX is to force the last-modified time of a file to be
updated, as the following example demonstrates.
% ls -l iecc.list
-rw------- 1 taylor
% touch iecc.list
% ls -l iecc.list
-rw------- 1 taylor
3843 Oct
6 18:02 iecc.list
3843 Oct 10 16:22 iecc.list
Because the touch command changes modification times of files, anything that sorts files
based on modification time will, of course, alter the position of that file when the file is altered
by touch.
1. Consider the following output:
4
% ls -t
mailing.lists
Cancelled.mail
RUMORS.18Sept
LISTS
newlists
iecc.list
News/
bin/
InfoWorld/
OWL/
Mail/
Archives/
src/
% touch iecc.list
% ls -t
iecc.list
RUMORS.18Sept
mailing.lists
LISTS
Cancelled.mail newlists
News/
bin/
InfoWorld/
OWL/
Mail/
Archives/
79
src/
You probably will not use touch for this purpose very often.
2. If you try to use the touch command on a file that doesn’t exist, the program
creates the file:
% ls
Archives/
LISTS
Cancelled.mail Mail/
InfoWorld/
News/
% touch new.file
% ls
Archives/
LISTS
Cancelled.mail Mail/
InfoWorld/
News/
% ls -l new.file
-rw-rw---- 1 taylor
OWL/
RUMORS.18Sept
bin/
iecc.list
mailing.lists
newlists
src/
OWL/
RUMORS.18Sept
bin/
iecc.list
mailing.lists
new.file
newlists
src/
0 Oct 10 16:28 new.file
The new file has zero bytes, as can be seen by the ls -l output. Notice that by
default the files are created with read and write permission for the user and anyone
in the user’s group. You learn in another hour how to determine, by using the
umask command, your own default permission for files.
You won’t need touch very often, but it’s valuable to know.
Task 4.10: Check Disk-Space Usage with du
One advantage that the DOS and Macintosh systems have over UNIX is they make
it easy to find out how much disk space you’re using and how much remains
available. On a Macintosh, viewing folders by size shows disk space used, and the top-right
corner of any Finder window shows available space. In DOS it’s even easier; both items are
listed at the end of the output from a DIR command:
C> DIR .BAT
Volume in drive C is MS-DOS_5
Volume Serial Number is 197A-A8D7
Directory of C:\
AUTOEXEC BAT
142 02-28-93
8:19p
CSH
BAT
36 12-22-92
3:01p
2 file(s)
178 bytes
5120000 bytes free
In this DOS example, you can see that the files listed take up 178 bytes, and that there are
5,120,000 bytes (about 5 megabytes, or 5MB) available on the hard drive.
Like a close-mouthed police informant, UNIX never volunteers any information, so you
need to learn two new commands. The du, disk usage, command is used to find out how much
disk space is used; the df, disk free, command is used to find out how much space is available.
4
Hour 4
80
1. The du command lists the size, in kilobytes, of all directories at or below the
current point in the file system.
% du
11
38
20
14
28
710
191
25
35
1627
./OWL
./.elm
./Archives
./InfoWorld/PIMS
./InfoWorld
./Mail
./News
./bin
./src
.
Notice that du went two levels deep to find the InfoWorld/PIMS subdirectory,
adding its size to the size indicated for the InfoWorld directory. At the very end, it
lists 1,627KB as the size of the dot directory—the current directory. As you know,
1,024KB kilobytes is a megabyte. Through division, you’ll find that the InfoWorld
directory is taking up 1.5MB of disk space.
2. If you are interested in only the grand total, you can use the -s flag to output just a
summary of the information.
% du -s
1627
.
Of course, you can look anywhere on the file system, but the more subdirectories
there are, the longer it takes.
3. Error messages with du are possible:
% du -s /etc
/etc/shadow: Permission denied
4417
/etc
In this example, one of the directories within the /etc directory has a permissions
set denying access:
% ls -ld /etc/shadow
drwx------ 2 root
512 Oct 10 16:34 /etc/shadow/
The du command summarizes disk usage only for the files it can read, so regardless
of the size of the shadow directory, I’d still have the 4,417 kilobytes size indicated.
4. Although by default du lists only the sizes of directories, it also computes the size of
all files. If you’re interested in that information, you can, by adding the -a flag,
have the program list it for all files.
% cd InfoWorld
% du -a
9
./PIM.review.Z
4
5
4
1
2
2
2
2
2
1
14
28
81
./Expert.opinion.Z
./PIMS/proposal.txt.Z
./PIMS/task1.txt.Z
./PIMS/task2.txt.Z
./PIMS/task3.txt.Z
./PIMS/task4.txt.Z
./PIMS/task5.txt.Z
./PIMS/task6.txt.Z
./PIMS/contact.info.Z
./PIMS
.
The problems of the -a flag for du are similar to those for the -R flag for ls. There
may be more files in a directory than you care to view.
5. The -a flag for listing all files overrides the -s flag for summarizing, but without
telling you it’s doing so. A preferable way would be for the program to note that
the two flags are incompatible, as many UNIX programs indicate, but that isn’t
how du works.
% du -s -a
9
./PIM.review.Z
5
./Expert.opinion.Z
4
./PIMS/proposal.txt.Z
1
./PIMS/task1.txt.Z
2
./PIMS/task2.txt.Z
2
./PIMS/task3.txt.Z
2
./PIMS/task4.txt.Z
2
./PIMS/task5.txt.Z
2
./PIMS/task6.txt.Z
1
./PIMS/contact.info.Z
28
.
6. The du command is an exception to the rule that multiple flags can be more
succinctly stated as a single multiletter flag. With ls, you’ll recall, -a -F -l could
be more easily typed as -aFl. The command du does not allow similar shorthand.
% du -sa
-sa: No such file or directory
UNIX is nothing if not varied. Some systems will accept du -as, and others
will not accept du -a -s. Try yours and see what does and doesn’t work.
JUST A MINUTE
It isn’t a problem that du does not allow multiletter flags, however,
because you do not use the -s and -a flags to du at the same time.
CAUTION
4
Hour 4
82
Task 4.11: Check Available Disk Space with df
Figuring out how much disk space is available on the overall UNIX system is difficult
for everyone except experts. The df command is used for this task, but it doesn’t
summarize its results—the user must add the column of numbers.
1. This is the system’s response to the df command:
% df
Filesystem
/dev/zd0a
/dev/zd8d
/dev/zd7d
/dev/zd3f
/dev/zd3g
/dev/zd2f
/dev/zd2g
/dev/zd1g
/dev/zd5c
/dev/zd0h
/dev/zd0g
kbytes
17259
185379
185379
385689
367635
385689
367635
301823
371507
236820
254987
used
14514
143995
12984
307148
232468
306189
207234
223027
314532
159641
36844
avail capacity
1019
93%
22846
86%
153857
8%
39971
88%
98403
70%
40931
88%
123637
63%
48613
82%
19824
94%
53497
75%
192644
16%
Mounted
/
/userf
/tmp
/users
/userc
/usere
/userb
/usera
/usr
/usr/src
/var
You end up with lots of information, but it isn't easy to add up quickly to find the
total space available. Nonetheless, the output offers quite a bit of information.
2. Because I know that my home directory is on the disk /users, I can simply look for
that directory in the rightmost column to find out that I’m using the hard disk
/dev/zd3f. I can see that there are 385,689KB on the disk, and 88 percent of the
disk is used, which means that 307,148KB are used and 39,971KB, or only about
38MB, are unused.
3. Some UNIX systems have relatively few separate computer disks hooked up,
making the df output more readable. The df output is explained in Figure 4.5.
% df
Filesystem
/dev/sd0a
/dev/sd2b
/dev/sd1a
Figure 4.5.
Understanding df
output.
4
kbytes
55735
187195
55688
used
37414
153569
43089
avail capacity
12748
75%
14907
91%
7031
86%
Mounted
/
/usr
/utils
83
You can add the columns to find that the system has a total of about 300MB of
disk space (55,735 + 187,195 + 55,688), of which 230MB are used. The remaining
space is therefore 33MB, or 16 percent of the total disk size.
Try using the du and df commands on your system to figure out how much disk space
is available on both the overall system and the disk you’re using for your home
directory. Then use du to identify how much space your files and directories are occupying.
Task 4.12: Shrink Big Files with the compress
Program
Now that you can figure out how much space you’re using with the files in your
directory, you’re ready to learn how to save space without removing any files. UNIX
has a built-in program—the compress program—that offers this capability.
1. In this simple example, the compress program is given a list of filenames and then
compresses each of the files, renaming them with a .Z suffix, which indicates that
they are compressed.
% ls -l LISTS
-rw------- 1 taylor
% compress LISTS
% ls -l LISTS.Z
-rw------- 1 taylor
106020 Oct 10 13:47 LISTS
44103 Oct 10 13:47 LISTS.Z
Compressing the LISTS file has reduced its size from 106KB to a little more than
44KB (a savings of almost 60 percent in disk space). If you expect to have large files
on your system that you won’t access very often, using the compress program can
save lots of disk space.
2. Using compress on bigger files can show even greater savings:
% ls -l huge.file
-rwxrwxrwx 1 root
% compress huge.file
% ls -l huge.file.Z
-rwxrwxrwx 1 taylor
3727360 Sep 27 14:03 huge.file
2121950 Sep 27 14:03 huge.file.Z
In this example, it took a powerful Sun computer with no other users exactly 20
seconds to compress huge.file. This single command was able to free over 1.5MB
of disk space. If you’re using a PC to run UNIX, or if you are on a system with
many users (which you can easily ascertain by using the w command), it might take
a significant amount of time to compress files.
4
Hour 4
84
3. To reverse the operation, use the companion command uncompress, and specify
either the current name of the file (that is, with the .Z suffix) or the name of the file
before it was compressed (that is, without the .Z suffix).
% uncompress LISTS
% ls -l LISTS
-rw------- 1 taylor
JUST A MINUTE
106020 Oct 10 13:47 LISTS
Why would you compress files? You would do so to save file space.
Before you use any of the compressed files, though, you must uncompress
them, so the compress utility is best used with large files you won’t need
for a while.
4. For information on how well the compress program shrunk your files, you can add
a -v flag to the program for verbose output:
% compress -v huge.file
huge.file: Compression: 43.15% -- replaced with huge.file.Z
Try using the compress program on some of the files in your directory, being careful
not to compress any files (particularly preference or dot files) that might be required
to run programs.
Summary
Most of this hour was spent learning about the powerful and complex ls command and its
many ways of listing files and directories. You also learned how to combine command flags
to reduce typing. You learned how to use the touch command to create new files and update
the modification time on older files, if needed. The hour continued with a discussion of how
to ascertain the amount of disk space you’re using and how much space is left, using the du
and df commands, respectively. Finally, you learned how the compress command can keep
you from running out of space by ensuring that infrequently used files are stored in the
minimum space needed.
Workshop
in the next hour.
Key Terms
4
access permission The set of accesses (read, write, and execute) allowed for each of the
three classes of users (owner, group, and everyone else) for each file or directory on the system.
85
block At its most fundamental, a block is like a sheet of information in the virtual notebook
that represents the disk: A disk is typically composed of many tens, or hundreds, of thousands
of blocks of information, each 512 bytes in size. You also might read the explanation of i-node
in the Glossary to learn more about how disks are structured in UNIX.
column-first order When you have a list of items that are listed in columns and span
multiple lines, column-first order is a sorting strategy in which items are sorted so that the
items are in alphabetical order down the first column and continuing at the top of the second
column, then the third column, and so on. The alternative strategy is row-first order.
permission strings The string that represents the access permissions.
row-first order In contrast to column-first order, this is when items are sorted in rows so
that the first item of each column in a row is in alphabetical order from left to right, then the
second line contains the next set of items, and so on.
Questions
1. Try using the du command on different directories to see how much disk space
each requires. If you encounter errors with file permissions, use ls -ld to list the
permissions of the directory in question.
2. Why would you want all the different types of sorting alternatives available with ls?
Can you think of situations in which each would be useful?
3. Use a combination of the ls -t and touch commands to create a few new files.
Then update their modification times so that in a most-recently-modified listing of
files, the first file you created shows up ahead of the second file you created.
4. Try using the du -s .. command from your home directory. Before you try it,
however, what do you think will happen?
5. Use df and bc or dc to figure out the amounts of disk space used and available on
your system.
6. Use the compress command to shrink a file in /tmp or your home directory. Use the
-v flag to learn how much the file was compressed, and then restore the file to its
original condition.
The next hour is a bit easier. It offers further explanation of the various information given
by the ls command and a discussion of file ownership, including how to change the owner
and group of any file or directory. You will learn about the chmod command, which can change
the specific set of permissions associated with any file or directory, and the umask command,
which can control the modes that new files are given upon creation.
4
Hour
87
5
Ownership and
Permissions
This hour focuses on teaching the basics of UNIX file permissions. Topics
include setting and modifying file permissions with chmod, analyzing file
permissions as shown by the ls -l command, and setting up default file
permissions with the umask command. Permission is only half the puzzle,
however, and you also learn about file ownership and group ownership, and how
to change either for any file or directory.
Goals for This Hour
■
■
■
■
■
Understand file permissions settings
Understand directory permissions settings
Modify file and directory permissions with chmod
Set new file permissions with chmod
Establish default file and directory permissions with umask
5
Hour 5
88
■ Identify the owner and group for any file or directory
■ Change the owner of a file or directory
■ Change the group of a file or directory
The preceding hour contained the first tutorial dealing with the permissions of a file or
directory using the -l option with ls. If you haven’t read that hour recently, it would help
to review the material. In this hour, you learn about another option to ls that tells UNIX to
show the group and owner of files or directories. Four more commands are introduced and
discussed in detail: chmod for changing the permissions of a file, umask for defining default
permissions, chown for changing ownership, and chgrp for changing the group of a file or
directory.
As you have seen in examples throughout the book, UNIX treats all directories as files; they
have their own size (independent of their contents), their own permissions strings, and more.
As a result, unless it’s an important difference, from here on I talk about files with the
intention of referring to files and directories both. Logic will confirm whether commands can
apply to both, or to files only, or to directories only. (For example, you can’t edit a directory
and you can’t store files inside other files.)
Task 5.1: Understand File Permissions Settings
In the last hour you learned a bit about how to interpret the information that ls offers
on file permissions when ls is used with the -l flag. Consider the following example.
% ls -l
total 403
drwx-----drwx------rw------drwx-----drwx-----drwx------rw------drwx------rw-------rw-rw-r--rw-rw---drwx------
2
3
1
2
2
2
1
2
1
1
1
2
taylor
taylor
taylor
taylor
taylor
taylor
taylor
taylor
taylor
taylor
taylor
taylor
512
512
106020
1024
512
512
4643
512
3843
280232
1031
512
Sep
Oct
Oct
Sep
Oct
Sep
Oct
Oct
Oct
Oct
Oct
Oct
30
1
10
30
6
30
10
10
10
10
7
10
10:38
08:23
13:47
10:50
09:36
10:51
14:01
19:09
16:22
16:22
15:44
19:09
Archives/
InfoWorld/
LISTS
Mail/
News/
OWL/
RUMORS.18Sept
bin/
iecc.list
mailing.lists
newlists
src/
The first item of information on each line is what is key here. You learned in the previous hour
that the first item is called the permissions string or, more succinctly, permissions. It also is
sometimes referred to as the mode or permissions mode of the file, a mnemonic that can be
valuable for remembering how to change permissions.
5
The permissions can be broken into four parts: type, owner, group, and other permissions.
The first character indicates the file type: d is a directory and - is a regular file. There are a
number of other types of files in UNIX, each indicated by the first letter of its permissions
string, as summarized in Table 5.1. You can safely ignore, however, any file that isn’t either
a regular file or directory.
89
Table 5.1. The ls file type indicators.
Letter
Indicated File Type
d
b
c
l
p
s
-
Directory
Block-type special file
Character-type special file
Symbolic link
Pipe
Socket
Regular file
The next nine letters in the permissions string are broken into three groups of three each—
representing the owner, group, and everyone else—as shown in Figure 5.1.
Figure 5.1.
Interpreting file
permissions.
To understand what the permissions actually mean to the computer system, remember that
UNIX treats everything as a file. If you install an application, it’s just like everything else, with
one exception: the system knows that an application is executable. A letter to your Mum is
a regular file, but if you were to tell UNIX that it was executable, the system would merrily
try to run it as a program (and fail).
There are three primary types of permission for files: read, write, and execute. Read
permission enables users to examine the contents of the file with a variety of different
programs, but they cannot alter, modify, or delete any information. They can copy the file
and then edit the new version, however.
Write permission is the next step up. Users with write access to a file can add information to
the file. If you have write permission and read permission for a file, you can edit the file—
the read permission enables you to view the contents, and the write permission lets you alter
them. With write permission only, you’d be able to add information to the file, but you
wouldn’t be able to view the contents of the file at any time. Admittedly, write-only
permission is unusual in UNIX, but you might see it for log files, which are files that track
activity on the system. Imagine if each time anyone logged in to your UNIX system the
computer recorded the fact, noting who logged in, where they logged in from, and the current
5
Hour 5
90
time and date. Armed with that information, you could ascertain who last logged in, who uses
dial-up phone lines, and who uses the computer the most. (In fact, there’s a UNIX command
that does just that. It’s called last.)
So far you’ve learned that you can have files with read-only permission, read-write
permission, and write-only permission. The third type of access permission is execute, noted
by ls with an x in the third slot of the permissions string.
% ls -l bin
total 57
-rwx------rwxrwx---rwx------rw-------rwx------
1
1
1
1
1
taylor
taylor
taylor
taylor
taylor
1507
32916
18567
334
3424
Aug
Oct
Sep
Oct
Sep
17
10
14
1
10
13:27
19:09
22:14
09:53
22:27
bounce.msg
calc
fixit
punt
rumor.mill.sh
1. Try listing the files in the directory /etc on your system, and see if you can identify
which are executable files or programs, which are directories, which are symbolic
links (denoted with an l as the first character of the permissions string; they’re files
that point to other files or directories that point to other directories), and which are
regular files.
2. Execute permission is slightly different from either read or write permission. If the
directory containing the file is in your search path (the value of the environment
variable PATH), any file that has execute permission is automatically started each
time that filename is entered, regardless of where you are in the file system.
% pwd
/users/taylor
% env PATH
/users/taylor/bin:/bin:/usr/bin:/usr/ucb:/usr/local:/usr/local/bin:
% ls -l bin/say.hi
-rwxrwx--- 1 taylor
9 Oct 11 13:32 bin/say.hi
% say.hi
hi
You can now see the importance of your search PATH. Without a search PATH, the
system wouldn’t be able to find any commands, and you’d be left with a barely
functional system. You can also see the purpose of checking the executable permission status. I’m going to jump ahead a bit to show you one use of the chmod
function so that you can see what happens if I remove the execute permission from
the say.hi program:
% chmod -x bin/say.hi
% ls -l bin/say.hi
-rw-rw---- 1 taylor
9 Oct 11 13:32 bin/say.hi
% say.hi
/users/taylor/bin/say.hi: Permission denied.
5
91
This time UNIX searched through my search path, found a file that matched the
name of the program I requested, and then ascertained that it wasn’t executable.
The resultant error message: Permission denied.
3. Now try entering say.hi on your computer system. You’ll get a different error
message, Command not found, which tells you that UNIX searched all the directories
in your search path but couldn’t find a match anywhere.
4. Check your PATH and find a directory that you can add files in. You’ll probably
have a bin directory in your home directory on the list, as I have /users/taylor/
bin in my search path. That’s a good place to add a file using the touch command:
% env PATH
/users/taylor/bin:/bin:/usr/bin:/usr/ucb:/usr/local:/usr/local/bin:
% touch bin/my.new.cmd
% ls -l bin
-rw-rw---- 1 taylor
0 Oct 11 15:07 my.new.cmd
5. Now try to actually execute the command by entering its name directly:
% my.new.cmd
/users/taylor/bin/my.new.cmd: Permission denied.
JUST A MINUTE
If you’re using the C Shell as your command interpreter, it probably won’t
find the new command you just created. This is because, to speed things
up, it keeps an internal table of where different commands are found in
your search path. You need to force the program to rebuild its table, and
you can do that with the simple command rehash. If, when you enter the
filename, you don’t get permission denied but instead see Command not
found, enter rehash and try again.
6. Finally, use chmod to add execute permission to the file, and try executing it one
more time.
% chmod +x bin/my.new.cmd
% ls -l bin/my.new.cmd
-rwxrw---- 1 taylor
% my.new.cmd
%
0 Oct 11 15:07 bin/my.new.cmd
Voila! You’ve created your first UNIX command, an achievement even though it
doesn’t do much. You can now see how the search path and the UNIX philosophy
of having applications be identical to regular files, except for the permission, can be
invaluable as you learn how to customize your environment.
Execute permission enables the user to run the file as if it were a program. Execute
permission is independent of other permissions granted—or denied—so it’s perfectly
feasible to have a program with read and execute permission, but no write permission. (After
all, you wouldn’t want others altering the program itself.) You also can have programs with
5
Hour 5
92
execute permission only. This means that users can run the application, but they can’t
examine it to see how it works or copy it. (Copying requires the ability to read the file.)
Though actual programs with execute-only permission work fine, a special
class of programs called shell scripts fail. Shell scripts act like a UNIX
command-line macro facility, which enables you to save easily a series of
commands in a file and then run them as a single program. To work,
however, the shell must be able to read the file and execute it, too, so
shell scripts always require both read and execute permissions.
JUST A MINUTE
There are clearly quite a few permutations on the three different permissions: read, write, and
execute. In practice, there are a few that occur most commonly, as listed in Table 5.2.
Table 5.2. The most common file permissions.
Permission
Meaning
---
No access is allowed
Read-only access
Read and execute access, for programs and shell scripts
Read and write access, for files
All access allowed, for programs
r-r-x
rwrwx
These permissions have different meanings when applied to directories, but
indicates that no one can access the file in question.
---
always
Interpretation of the following few examples should help:
-rw-------rw-rw-r--rw-rw----rwxr-x---
1
1
1
1
taylor
taylor
taylor
taylor
3843
280232
1031
64
Oct 10 16:22 iecc.list
Oct 10 16:22 mailing.lists
Oct 7 15:44 newlists
Oct 9 09:31 the.script
The first file, iecc.list, has read and write permission for the owner (taylor) and is off-limits
to all other users. The file mailing.lists offers similar access to the file owner (taylor) and
to the group but offers read-only access to everyone else on the system. The third file,
newlists, provides read and write access to both the file owner and group, but no access to
anyone not in the group.
5
93
The fourth file on the list, the.script, is a program that can be run by both the owner and
group members, read (or copied) by both the owner and group, and written (altered) by the
owner. In practice, this probably would be a shell script, as described earlier, and these
permissions would enable the owner (taylor) to use an editor to modify the commands
therein. Other members of the group could read and use the shell script but would be denied
access to change it.
Task 5.2: Directory Permissions Settings
Directories are similar to files in how you interpret the permissions strings. The
differences occur because of the unique purpose of directories, namely to store other
files or directories. I always think of directories as bins or boxes. You can examine the box
itself, or you can look at what’s inside.
In many ways, UNIX treats directories simply as files in the file system, where the content
of the file is a list of the files and directories stored within, rather than a letter, program, or
shopping list.
The difference, of course, is that when you operate with directories, you’re operating both
with the directory itself, and, implicitly, with its contents. By analogy, when you fiddle with
a box full of toys, you’re not altering just the state of the box itself, but also potentially the
toys within.
There are three permissions possible for a directory, just as for a file: read, write, and execute.
The easiest is write permission. If a directory has write permission enabled, you can add new
items and remove items from the directory. It’s like owning the box; you can do what you’d
like with the toys inside.
The interaction between read and execute permissions with a directory is confusing. There
are two types of operations you perform on a directory: listing the contents of the directory
(usually with ls) and examining specific, known files within the directory.
1. Start by listing a directory, using the -d flag:
% ls -ld testme
dr-x------ 2 taylor
% ls -l testme
total 0
-rw-rw---- 1 taylor
% ls -l testme/file
-rw-rw---- 1 taylor
512 Oct 11 17:03 testme/
0 Oct 11 17:03 file
0 Oct 11 17:03 testme/file
For a directory with both read and execute permissions, you can see that it’s easy to
list the directory, find out the files therein, and list specific files within the directory.
5
Hour 5
94
2. Read permission on a directory enables you to read the “table of contents” of the
directory but, by itself, does not allow you to examine any of the files therein. By
itself, read permission is rather bizarre:
% ls -ld testme
dr-------- 2 taylor
% ls -l testme
testme/file not found
total 0
% ls -l testme/file
testme/file not found
512 Oct 11 17:03 testme/
Notice that the system indicated the name of the file contained in the testme
directory. When I tried to list the file explicitly, however, the system couldn’t find
the file.
3. Compare this with the situation when you have execute permission—which enables
you to examine the files within the directory—but you don’t have read permission,
and you are prevented from viewing the table of contents of the directory itself:
% ls -ld testme
d--x------ 2 taylor
% ls -l testme
testme unreadable
% ls -l testme/file
-rw-rw---- 1 taylor
512 Oct 11 17:03 testme/
0 Oct 11 17:03 testme/file
With execute-only permission, you can set up directories so that people who know
the names of files contained in the directories can access those files, but people
without that knowledge cannot list the directory to learn the filenames.
4. I’ve actually never seen anyone have a directory in UNIX with execute-only
permission, and certainly you would never expect to see one set to read-only. It
would be nice if UNIX would warn you if you set a directory to have one permission and not the other. However, UNIX won’t do that. So, remember for directories always to be sure that you have both read and execute permission set. Table 5.3
summarizes the most common directory permissions.
Table 5.3. The most common directory permissions.
Permission
Meaning
---
No access allowed to directory
Read-only access, no modification allowed
All access allowed
r-x
rwx
5
95
5. One interesting permutation of directory permissions is for a directory that’s writeonly. Unfortunately, the write-only permission doesn’t do what you’d hope, that is,
enable people to add files to the directory without being able to see what the
directory already contains. Instead, it’s functionally identical to having it set for no
access permission at all.
At the beginning of this hour, I used ls to list various files and directories in my
home directory:
% ls -l
total 403
drwx-----drwx------rw------drwx-----drwx-----drwx------rw------drwx------rw-------rw-rw-r--rw-rw---drwx------
2
3
1
2
2
2
1
2
1
1
1
2
taylor
taylor
taylor
taylor
taylor
taylor
taylor
taylor
taylor
taylor
taylor
taylor
512
512
106020
1024
512
512
4643
512
3843
280232
1031
512
Sep
Oct
Oct
Sep
Oct
Sep
Oct
Oct
Oct
Oct
Oct
Oct
30
1
10
30
6
30
10
10
10
10
7
10
10:38
08:23
13:47
10:50
09:36
10:51
14:01
19:09
16:22
16:22
15:44
19:09
Archives/
InfoWorld/
LISTS
Mail/
News/
OWL/
RUMORS.18Sept
bin/
iecc.list
mailing.lists
newlists
src/
Now you can see that all my directories are set so that I have list, examine, and
modify (read, execute, and write, respectively) capability for myself, and no access is
allowed for anyone else.
6. The very top-level directory is more interesting, with a variety of different directory
owners and permissions:
% ls -l /
-rw-r--r-drwxr-xr-x
-r--r--r-drwxr-xr-x
drwxr-xr-x
lrwxr-xr-x
drwxrwxrwx
drwxr-xr-x
drwxr-xr-x
drwxr-xr-x
drwxr-xr-x
drwxr-xr-x
drwxr-xr-x
1
4
1
6
2
1
65
753
317
626
534
34
5
root
root
root
root
root
root
root
root
root
root
root
root
root
61440
23552
686753
3072
8192
7
12800
14848
13312
13312
10752
1024
1024
Nov
Sep
Aug
Oct
Apr
Jul
Oct
Oct
Oct
Oct
Sep
Oct
Oct
29
27
27
11
12
28
11
5
5
8
30
1
1
1991
11:31
21:58
16:30
1991
1988
17:33
10:07
10:17
13:02
13:06
09:10
09:20
boot
dev
dynix
etc
lost+found
sys -> usr/sys
tmp
usera
userb
userc
users
usr
var
Clearly, this machine has a lot of users. Notice that the link count for usera, userb,
userc, and users are each in the hundreds. The dev directory has read and execute
permission for everyone and write permission for the owner (root). Indeed, all the
directories at this level are identical except for tmp, which has read, write, and
execute permission for all users on the system.
5
Hour 5
96
7. Did you notice the listing for the sys directory buried in that output?
lrwxr-xr-x
1 root
7 Jul 28
1988 sys -> usr/sys
From the information in Table 5.1, you know that the first letter of the permissions string being an l means that the directory is a symbolic link. The filename
shows just the specifics of the link, indicating that sys points to the directory usr/
sys. In fact, if you count the number of letters in the name usr/sys, you’ll find that
it exactly matches the size of the sys link entry, too.
8. Try using ls -l / yourself. You should be able to understand the permissions of
any file or directory that you encounter.
Permissions of files and directories will prove easier as you work with UNIX more.
Task 5.3: Modify File and Directory Permissions
with chmod
Now that you can list directory permissions and understand what they mean, how
about learning a UNIX command that lets you change them to meet your needs?
You’ve already had a sneak preview of the command: chmod. The mnemonic is “change
mode,” and it derives from early UNIX folk talking about permission modes of files. You can
remember it by thinking of it as a shortened form of change permission modes.
To sound like a UNIX expert, pronounce chmod as “ch-mod,” “ch” like the
beginning of child, and “mod” to rhyme with cod.
JUST A MINUTE
The chmod command enables you to specify permissions in two different ways: symbolically
or numerically. Symbolic notation is most commonly used to modify existing permissions,
whereas numeric format always replaces any existing permission with the new value specified.
In this task, you learn about symbolic notation, and the next task focuses on the powerful
numeric format.
Symbolic notation for chmod is a bit like having a menu of different choices, enabling you to
pick the combination that best fits your requirements. Figure 5.2 shows the two menus.
Figure 5.2.
The menu of symbolic
chmod values.
5
97
The command chmod is like a smorgasbord where you can choose any combination of items
from either the first or last boxes, and stick your choice from the center box between them.
For example, if you wanted to add write permission to the file test for everyone in your group,
you would, working backwards from that description, choose g for group, + for add, and w
for write. The finished UNIX command would be chmod g+w test.
If you decided to take away read and execute permission for everyone not in your group, you
could use chmod o-rx test to accomplish the task.
1. Turn to your computer and, using touch and ls, try changing permissions and see
what happens. I’ll do the same:
% touch test
% ls -l test
-rw-rw---- 1 taylor
0 Oct 11 18:29 test
2. The first modification I want to make is that people in my group should be able to
read the file. Because I don’t really want them altering it, I’ll rescind write permission for group members:
% chmod g-w test
% ls -l test
-rw-r----- 1 taylor
0 Oct 11 18:29 test
3. But then my boss reminds me that everyone in the group should have all access
permissions for everyone in that group. Okay, I’ll do so.
% chmod g+wx test
% ls -l test
-rw-rwx--- 1 taylor
5
0 Oct 11 18:29 test
I also could have done that with chmod g=rwx, of course.
4. Wait a second. This test file is just for my own use, and nobody in my group
should be looking at it anyway. I’ll change it back.
% chmod g-rwx test
% ls -l test
-rw------- 1 taylor
0 Oct 11 18:29 test
Great. Now the file is set so that I can read and write it, but nobody else can touch
it, read it, modify it, or anything else.
5. If I relented a bit, I could easily add, with one last chmod command, read-only
permission for everyone:
% chmod a+r test
% ls -l test
-rw-r--r-- 1 taylor
0 Oct 11 18:29 test
Hour 5
98
Permissions in UNIX are based on a concentric access model from Multics. (In Hour 1,
you learned that the name UNIX is also a pun on Multics.) Figure 5.3 illustrates this
concept.
Figure 5.3.
The concentric circles of
access.
As a result, it’s incredibly rare to see a file where the owner doesn’t have the most access to
a file. It’d be like buying a car and letting everyone but yourself drive it—rather silly.
Similarly, members of the group are given better or equal permission to everyone else on the
machine. You would never see r--r--rwx as a permissions string.
Experiment a bit more with the various combinations possible with the chmod symbolic
notation. How would you change permission on a directory to enable all users to use ls to
examine it but to deny them the ability to add or remove files? How about adding write access
for the owner but removing it for everyone else?
Task 5.4: Set New File Permissions with chmod
The second form of input that chmod accepts is absolute numeric values for
permissions. Before you can learn how to use this notation, you have to learn a bit
about different numbering systems first.
The numbering system you’re familiar with, the one you use to balance your checkbook and
check the receipt from the market, is decimal, or base 10. This means that each digit—from
right to left—has the value of the digit raised by a power of 10, based on the digit’s location
in the number. Figure 5.4 shows what the number 5,783 is in decimal.
You can see that in a base-10 numbering system, the value of a number is the sum of the value
of each digit multiplied by the numeric base raised to the nth power. The n is the number
of spaces the digit is away from the rightmost digit. That is, in the number 5,783, you know
that the 7 is worth more than just 7, because it’s two spaces away from the rightmost digit
5
99
(the 3). Therefore, its value is the numeric base (10) raised to the nth power, where n is 2 (it’s
two spaces away). Ten to the second power equals 100 (102 = 100), and when you multiply
that by 7, sure enough, you find that the 7 is worth 700 in this number.
Figure 5.4.
Interpreting decimal
numbers.
What does all this have to do with the chmod command? At its most fundamental, UNIX
permissions are a series of on/off switches. Does the group have write permission? One equals
yes, zero equals no. Each digit in a decimal system can have 10 different values. A binary
system is one in which each digit can have only two values: on or off, yes or no. Therefore,
you can easily and uniquely describe any permissions string as a series of zeroes and ones—
as a binary number. Figure 5.5 demonstrates.
Figure 5.5.
Permissions as binary
numbers.
5
The convention is that if a letter is present, the binary digit is a 1—that permission is
permitted—and if no letter is present, the digit is a zero. Thus, r-xr----- can be described
as 101100000, and r--r--r-- can be described in binary as 100100100.
You’ve already learned that the nine-character permissions string is really just a threecharacter permissions string duplicated thrice for the three different types of user (the owner,
group, and everyone else). That means that you can focus on learning how to translate a single
tri-character permissions substring into binary and extrapolate for more than one permission.
Table 5.3 lists all possible permissions and their binary equivalents.
Hour 5
100
Table 5.3. Permissions and binary equivalents.
Permissions
String
Binary
Equivalent
---
000
001
010
011
100
101
110
111
--x
-w-wx
r-r-x
rwrwx
Knowing how to interpret decimal numbers using the rather complex formula presented
earlier, you should not be surprised that the decimal equivalent of any binary number can be
obtained by the same technique. Figure 5.6 shows how, with the binary equivalent of the
r-x permission.
Figure 5.6.
Expressing r-x as a single
digit.
If r-x is equal to 5, it stands to reason that each of the possible three-character permissions
has a single-digit equivalent, and Table 5.4 expands Table 5.3 to include the single-digit
equivalents.
5
101
Table 5.4. Permissions and numeric equivalents.
Permissions
String
Binary
Equivalent
Decimal
Equivalent
---
000
001
010
011
100
101
110
111
0
1
2
3
4
5
6
7
--x
-w-wx
r-r-x
rwrwx
The value of having a single digit to describe any of the seven different permission states
should be obvious. Using only three digits, you now can fully express any possible
combination of permissions for any file or directory in UNIX—one digit for the owner
permission, one for group, and one for everyone else. Figure 5.7 shows how to take a full
permissions string and translate it into its three-digit numeric equivalent.
Figure 5.7.
Translating a full
permissions string into its
numeric equivalent.
5
From this illustration, you can see how the permissions string rw-r----- (read and write
permission for the owner, read permission for the group, and no access allowed for everyone
else) is exactly equivalent to the numeric string 640.
1. Try to create numeric strings on your own, using Table 5.4 to help. Turn to your
computer and use ls to display some listings. Break each permissions string into
three groups of three letters, and figure out the numeric equivalents. Here are some
examples from the ls –C –F listing of my home directory:
drwx------
2 taylor
512 Sep 30 10:38 Archives/
For Archives/, the equivalent numeric permission is 700.
Hour 5
102
-rw-------
1 taylor
106020 Oct 10 13:47 LISTS
For LISTS, the equivalent numeric permission is 600.
-rw-rw-r--
1 taylor
280232 Oct 10 16:22 mailing.lists
For mailing.lists, the equivalent numeric permission is 664.
-rw-rw----
1 taylor
1031 Oct
7 15:44 newlists
For newlists, the equivalent numeric permission is 660.
There’s one last step required before you can try using the numeric permissions strings
with chmod. You need to be able to work backwards to determine a permission that
you’d like to set, and figure out the numeric equivalent for that permission.
Task 5.5: Calculating Numeric Permissions Strings
For example, if you wanted to have a directory set so that you have all access, people
in your group can look at the contents but not modify anything, and everyone else
is shut out, how would you do it?
All permissions for yourself means you want read+write+execute for owner (or numeric
permission 7); read and listing permission for others in the group means read+execute for
group (numeric permission 5); and no permission for everyone else, numeric permission 0.
Put the three together and you have the answer, 750.
That’s the trick of working with chmod in numeric mode. You specify the absolute permissions
you want as a three-digit number, and the system sets the permissions on the file or directory
appropriately.
The absolute concept is important with this form of chmod. You cannot use the chmod numeric
form to add or remove permissions from a file or directory. It is usable only for reassigning
the permissions string of a file or directory.
The good news is that, as you learned earlier in this hour, there is a relatively small number
of commonly-used file permissions, summarized in Table 5.5.
Table 5.5. Common permissions and their numeric equivalents.
Permission
Numeric
Used With
---------
000
400
444
600
644
664
All types
Files
Files
Files
Files
Files
r-------r--r--r-rw------rw-r--r-rw-rw-r--
5
Permission
Numeric
Used With
rw-rw-rw-
666
700
750
755
Files
Programs and directories
rwx-----rwxr-x--rwxr-xr-x
103
1. Turn to your computer and try using the numeric mode of chmod, along with ls, to
display the actual permissions to learn for yourself how this works.
% touch example
% ls -l example
-rw-rw---- 1 taylor
0 Oct 12 10:16 example
By default, files are created in my directory with mode 660.
2. To take away read and write permission for people in my group, I’d replace the 660
permission with what numeric permissions string? I’d use 600:
% chmod 600 example
% ls -l example
-rw------- 1 taylor
3. What if I change my mind and want to open the file up for everyone to read or
write? I’d use 666:
% chmod 666 example
% ls -l example
-rw-rw-rw- 1 taylor
4. Finally, pretend that the example is actually a directory. What numeric mode
would I specify to enable everyone to use ls in the directory and enable only the
owner to add or delete files? I’d use 755:
% chmod 755 example
% ls -l example
-rwxr-xr-x 1 taylor
You’ve looked at both the numeric mode and the symbolic mode for defining
permissions. Having learned both, which do you prefer?
JUST A MINUTE
Somehow I’ve never gotten the hang of symbolic mode, so I almost
always use the numeric mode for chmod. The only exception is when I
want to add or delete simple permissions. Then, I use something like chmod
+r test to add read permission. Part of the problem is that I don’t think of
the user of the file but rather the owner, and specifying o+r causes chmod
5
104
Hour 5
to change permissions for others. It’s important, therefore, that you
remember that files have users so you remember u for user, and that
everyone not in the group is other so you remember o. Otherwise, learn
the numeric shortcut!
File permissions and modes are one of the most complex aspects of UNIX. You can tell—
it’s taken two hours to explain it fully. It’s very important that you spend the time really to
understand how the permissions strings relate to directory permissions, how to read the
output of ls, and how to change modes using both styles of the chmod command. It’ll be time
well spent.
Task 5.6: Establish Default File and Directory
Permissions with the umask Command
When I’ve created files, they’ve had read+write permissions for the owner and group,
but no access allowed for anyone else. When you create files on your system, you
might find that the default permissions are different.
The controlling variable behind the default permissions is called the file creation mask, or
umask for short.
Inexplicably, umask doesn’t always list its value as a three-digit number, but you can find its
value in the same way you figured out the numeric permissions strings for chmod. For example,
when I enter umask, the system indicates that my umask setting is 07. A leading zero has been
dropped, so the actual value is 007, a value that British MI6 could no doubt appreciate!
But 007 doesn’t mean that the default file is created with read+write+execute for everyone
else and no permissions for the owner or group. It means quite the opposite, literally.
The umask command is a filter through which permissions are pushed to ascertain what
remains. Figure 5.8 demonstrates how this works.
Think of your mask as a series of boxes: if the value is true, the information can’t exude
through the box. If the value is false, it can. Your mask is therefore the direct opposite to how
you want your permissions to be set. In Figure 5.8, I want to have 770 as the default
permission for any new file or directory I create, so I want to specify the exact opposite of that,
007. Sure enough, with this umask value, when I create new files, the default permission allows
read and write access to the owner and group, but no access to anyone else.
5
105
Figure 5.8.
Interpreting the umask
value.
Things are a bit trickier than that. You’ve probably already asked yourself, “Why, if I have
007 as my mask (which results in 770 as the default permissions), do my files have 660 as the
actual default permission?”
The reason is that UNIX tries to be smart about the execute permission setting. If I create
a directory, UNIX knows that execute permission is important, and so it grants it. However,
for some files (particularly text files), execute permission doesn’t make sense, so UNIX
actually masks it out internally.
Another way to look at this is that any time you create a file containing information, the
original mask that the system uses to compare against your umask is not 777 (not rwxrwxrwx,
to put it another way), but rather 666 (rw-rw-rw-), in recognition of the unlikelihood that
you’ll want to execute the new file.
The good news is that you now know an easy way to set the execute permission for a file if
the system gets it wrong: chmod +x filename does the trick.
5
Hour 5
106
1. Turn to your computer and check your umask setting, then alternate between
changing its values and creating new files with touch:
% umask
7
% touch test.07
% ls -l test.07
-rw-rw---- 1 taylor
0 Oct 12 14:38 test.07
2. To change the value of your umask, add the numeric value of the desired mask to
the command line:
% umask 077
This changes my umask value from 007 (------rwx) to 077 (---rwxrwx). Before you
look at the following listing, what would you expect this modification to mean?
Remember, you should read it as the exact opposite of how you want the default
permissions.
% touch test.077
% ls -l test.077
-rw------- 1 taylor
0 Oct 12 14:38 test.077
Is that what you expected?
3. What would you do if you wanted to have the default permission keep files private
to just the owner and make them read-only?
You can work through this problem in reverse. If you want r-x------ as the default
permission (since the system takes care of whether execute permission is needed,
based on file type), write down the opposite permission, which is -w-rwxrwx.
Translate that to a binary number, 010 111 111, and then to a three-digit value,
277 (010=2, 111=7, 111=7). That’s the answer. The value 277 is the correct umask
value to ensure that files you create are read-only for yourself and off-limits to
everyone else.
% umask 277
% touch test.277
% ls -l test.277
-r-------- 1 taylor
0 Oct 12 14:39 test.277
4. What if you wanted to have files created with the default permission being readonly for everyone, read-write for the group, but read-only for the owner? Again,
work backwards. The desired permission is r-xrwxr-x, so create the opposite value
(-w-----w-), translate it into binary (010 000 010), and then translate that into a
three-digit value: 202 (010=2, 000=0, 010=2).
5
107
As a rule of thumb, it’s best to leave the execute permission enabled when
building umask values so the system doesn’t err when creating directories.
JUST A MINUTE
The umask is something set once and left alone. If you’ve tried various experiments
on your computer, remember to restore your umask back to a sensible value to avoid
future problems (though each time you log in to the system it’s reset to your default value).
In the next hour, you learn how to use the mkdir command to create new directories, and you
see how the umask value affects default directory access permissions.
Task 5.7: Identify Owner and Group for Any File or
Directory
One of the many items of information that the ls command displays when used with
the -l flag is the owner of the file or directory. So far, all the files and directories in
your home directory have been owned by you, with the probable exception of the “..”
directory, which is owned by whomever owns the directory above your home.
In other words, when you enter ls
every file in the listing.
-l,
you should see your account name as the owner for
If you’re collaborating with another user, however, there might well be times when you’ll
want to change the owner of a file or directory once you’ve created and modified it. The first
step in accomplishing this is to identify the owner and group.
Identifying the owner is easy; ls lists that by default. But how do you identify the group of
which the file or directory is a part?
1. The ls command can show the group membership of any file or directory by the
addition of a new command flag, -g. By itself, -g doesn’t alter the output of ls, but
when used with the -l flag, it adds a column of information to the listing. Try it on
your system. Here is an example:
% ls -lg /tmp
-rw-r--r-- 1
drwxr-xr-x 2
-rw------- 1
-rw------- 1
-rw------- 1
-rw-r----- 1
root
shakes
meademd
dessy
steen
jsmith
root
root
com435
stuprsac
utech
utech
0
512
0
1191
1
258908
Oct
Oct
Oct
Oct
Oct
Oct
12
12
12
12
12
12
14:52
07:23
14:46
14:57
10:28
12:37
sh145
shakes/
snd.12
snd.15
snd.17
sol2
5
Hour 5
108
On many System V-based systems, the output of ls -l always shows user
and group. The -g flag actually turns off this display!
JUST A MINUTE
Both owners and groups vary for each of the files and directories in this small
listing. Notice that files can have different owners while having the same group.
(There are two examples here: sh145 and the shakes directory, and snd.17 and
sol2.)
2. Directories that have a wide variety of owners are the directories above your own
home directory and the tmp directory, as you can see in instruction 1. Examine
both on your system and identify both the owner and group of all files. For files in
the same group you’re in (with the id command, you can find which group or
groups you are in) but not owned by you, you’ll need to check which of the three
permission values to identify your own access privileges?
Files and directories have both owners and groups, although the group is ultimately
less important than the owner, particularly where permissions and access are
involved.
Task 5.8: Change the Owner of a File or Directory
Now that you can ascertain the ownership of a file or directory, it’s time to learn
about the chown command. This command lets you change the ownership of
whatever you specify.
CAUTION
Before you go any further, however, a stern warning: once you’ve
changed the ownership of a file, you cannot restore it to yourself. Only the
owner of a file can give away its ownership, so don’t use the chown
command unless you’re absolutely positive you want to!
1. The format for changing the ownership of a file is to specify the new owner and
then list the files or directory you are giving away:
% ls -l test
-rwxrwxrwx 1 taylor
% chown root test
% ls -l test
-rwxrwxrwx 1 root
5
0 Oct 12 15:17 mytest
This would change the ownership of the file test from me to the user root on the
system.
109
2. If I now try to change the ownership back, it fails:
% chown taylor test
chown: test: Not owner
Most modern UNIX systems prevent users from changing the ownership of a file
due to the inherent dangers. If you try chown, and it returns Command not found or
Permission denied , that means you’re barred from making any file ownership
changes.
3. On one of the systems I use, chown always reports Not owner when I try to change a
file regardless of whether I really am the owner or not:
% ls -l mytest
-rwxrwxrwx 1 taylor
% chown root mytest
chown: mytest: Not owner
This is needlessly confusing—a message like “you’re not allowed to change file
ownership” would be better. But, alas, like so much of UNIX, it’s up to the user to
figure out what’s going on.
To change the ownership of a file or directory, you can use the chown command if you
have the appropriate access on your system. It’s like a huge supertanker, though; you
can’t change course once underway, so be cautious!
Task 5.9: Change the Group of a File or Directory
Changing the group membership of a file or directory is quite analogous to the steps
required for changing file ownership. Almost all UNIX systems enable users to use
the chgrp command to accomplish this task.
1. Usage of chgrp is almost identical to chown, too. Specify the name of the group,
followed by the list of files or directories to reassign:
% ls -lg
-rwxrwxrwx 1 taylor
% chgrp ftp mytest
% ls -lg
-rwxrwxrwx 1 taylor
ci
ftp
The caveat on this command, however, is that you must be a member of the group
you’re assigning for the file, or it fails:
% ls -lg
-rwxrwxrwx 1 taylor
ftp
% chgrp root mytest
chgrp: You are not a member of the root group
5
Hour 5
110
Portions of UNIX are well thought out and offer innovative approaches to common
computer problems. File groups and file ownership aren’t examples of this, unfortunately. The majority of UNIX users tend to be members of only one group, so they cannot
change the group membership or ownership of any file or directory on the system. Instead,
users seem to just use chmod to allow full access to files; then they encourage colleagues to copy
the files desired, or they simply allow everyone access.
Unlike the other commands you’ve learned in this book, chown might be one you will not use.
It’s entirely possible that you’ll never need to change the ownership or group membership
of any file or directory.
Summary
In this hour, you learned the basics of UNIX file permissions, including how to set and
modify file permissions with chmod and how to analyze file permissions as shown by the
ls -l command. You also learned about translating between numeric bases (binary and
decimal) and how to convert permissions strings into numeric values. Both are foundations
for the umask command, which you learned to interpret and alter as desired. Permission is only
half the puzzle, however, so you also learned about file ownership, group ownership, and how
to change either for any file or directory.
Workshop
in the next hour.
Key Terms
file creation mask When files are created in UNIX, they inherit a default set of access
permissions. These defaults are under the control of the user and are known as the file creation
mask.
mode A shorthand way of saying permissions mode.
permissions mode The set of accesses (read, write, and execute) allowed for each of the
This is a synonym for access permission.
shell script A collection of shell commands in a file.
5
111
Questions
1. In what situations might the following file permissions be useful?
r--rw-r--
r--r--rw-
rw--w--w-
-w--w--w-
rwxr-xr-x
r-x--x--x
2. Translate the six file permissions strings in instruction 1 into their binary and
numeric equivalents.
3. Explain what the following umask values would make the default permissions for
newly created files:
007
077
777
111
222
733
272
544
754
4. Count the number of groups that are represented by group membership of files in
the tmp directory on your system. Use id to see if you’re a member of any of them.
5. Which of the following directories could you modify, if the id command listed the
following information? Which could you view using the ls command?
% id
uid=19(smith)
% ls -lgF
drw-r--r-- 2
drwxr-xr-x 2
drw------- 2
drwxr-x--- 3
drwx------ 3
gid=50(users) groups=50(users)
root
shakes
meademd
smith
jin
users
root
com435
users
users
512
512
1024
512
512
Oct
Oct
Oct
Oct
Oct
12
12
12
12
12
14:52
07:23
14:46
12:37
12:37
sh/
shakes/
tmp/
viewer/
Zot!/
In the next hour, you learn the various UNIX file-manipulation commands, including how
to copy files, how to move them to new directories, and how to create new directories. You
also learn how to remove files and directories as well as about the dangers of file removal on
UNIX.
5
Hour
113
6
Creating, Moving,
Renaming, and
Deleting Files and
Directories
In this hour, you learn the basic UNIX file-manipulation commands. These
commands will explain how to create directories with mkdir, remove directories
with rmdir, use cp and mv to move files about in the file system, and use rm to
remove files. The rm command has its dangers: you learn that there isn’t an
“unremove” command in UNIX and how to circumvent the possible dangers
that lurk in the program.
Goals for This Hour
■ Create new directories using mkdir
■ Copy files to new locations using cp
5
6
Hour 6
114
■
■
■
■
■
Move files to new locations using mv
Rename files using mv
Remove directories using rmdir
Remove files using rm
Minimize the danger of using the rm command
This hour introduces several tremendously powerful commands that enable you to create a
custom file-system hierarchy (or wreak unintentional havoc on your files). As you learn these
commands, you also learn hints and ideas on how to best use the UNIX file system to keep
your files neat and organized. These simple UNIX commands, all new in this hour, are found
not only in all variants of UNIX, both BSD-based and System V-based, but they also can be
brought onto DOS through utilities such as the MKS Toolkit from Mortice-Kern Systems.
Task 6.1: Creating New Directories Using mkdir
One important aspect of UNIX that has been emphasized continually in this book
is that the UNIX file system is hierarchical. The UNIX file system includes
directories containing files and directories, each of which can contain both files and
directories. Your own home directory, however, probably doesn’t contain any directories
(except “.” and “..”, of course), which prevents you from exploiting what I call the virtual file
cabinet of the file system.
The command for creating directories is actually one of the least complex and most
mnemonic (for UNIX, anyway) in this book: mkdir, called “make directory.”
Pronounce the mkdir command as “make dir.”
JUST A MINUTE
1. Turn to your computer, move to your home directory, and examine the files and
directories there. Here’s an example:
% cd
% ls
Archives/
InfoWorld/
LISTS
Mail/
News/
6
OWL/
PubAccessLists.Z
bin/
educ
mailing.lists.bitnet.Z
rumors.26Oct.Z
rumors.5Nov.Z
src/
115
2. To create a directory, you need to specify what you’d like to name the directory
and where you’d like to locate it in the file system (the default location is your
current working directory):
% mkdir NEWDIR
% ls
Archives/
InfoWorld/
LISTS
Mail/
NEWDIR/
News/
OWL/
PubAccessLists.Z
bin/
educ
rumors.26Oct.Z
rumors.5Nov.Z
src/
3. That’s all there is to it. You’ve created your first UNIX directory, and you can now
list it with ls to see what it looks like:
% ls -ld NEWDIR
drwxrwx--- 2 taylor
% ls -la NEWDIR
total 2
drwxrwx--- 2 taylor
24 Nov
24 Nov
1024 Nov
5 10:48 NEWDIR/
5 10:48 ./
5 10:48 ../
Not surprisingly, the directory is empty other than the two default entries of “.”
(the directory itself) and “..” (the parent directory, your home directory).
4. Look closely at the permissions of the directory. Remember that the permissions
are a result of your umask setting. As you learned in the previous hour, changing the
umask setting changes the default directory permissions. Then, when you create a
new directory, the new permissions will be in place:
% umask
07
% umask 0
% mkdir NEWDIR2
% ls -ld NEWDIR2
drwxrwxrwx 2 taylor
% umask 222
% mkdir NEWDIR3
% ls -ld NEWDIR3
dr-xr-xr-x 2 taylor
5
24 Nov
5 10:53 NEWDIR2/
24 Nov
5 10:54 NEWDIR3/
5. What happens if you try to create a directory with a name that has already been
used?
% mkdir NEWDIR
mkdir: NEWDIR: File exists
6. To create a directory other than your current location, prefix the new directory
name with a location:
% mkdir /tmp/testme
% ls -l /tmp
-rwx------ 1 zhongqi
-rw------- 1 xujia
-rw-r--r-- 1 beast
22724 Nov
95594 Nov
572 Nov
4 21:33 /tmp/a.out*
4 23:10 /tmp/active.10122
5 05:59 /tmp/anon1
6
Hour 6
116
-rw-rw----rw-------rwx------rwx------rw-r--r--rw------drwxrwx---rw-r--r--
1
1
1
1
1
1
2
1
root
qsc
steen
techman
root
zhongqi
taylor
aru
0
0
24953
3711
997536
163579
24
90
Nov
Nov
Nov
Nov
Nov
Nov
Nov
Nov
5
5
5
5
5
4
5
5
10:30
00:18
10:40
10:45
10:58
20:16
10:56
02:55
/tmp/bar.report
/tmp/lh013813
/tmp/mbox.steen*
/tmp/mbox.techman*
/tmp/quotas
/tmp/sp500.1
testme/
/tmp/trouble21972
Like other basic UNIX utilities, mkdir has no command arguments, so it is quite easy
to use. There are two things to keep in mind: You must have write permission to the
current directory if you’re creating a new directory, and you should ensure that the name of
the directory is not the same as (or, to avoid confusion, similar to) a directory name that
already exists.
Task 6.2: Copying Files to New Locations Using cp
One of the most basic operations in any system is moving files, the modern-office
computer equivalent of paper shuffling. On a computer, moving files is a simple
matter of using one or two commands: you can move a file to a different location, or you can
create a copy of the file and move the copy to a different location.
The Macintosh has an interesting strategy for differentiating between moving and copying.
If you drag a file to another location that’s on the same device (a hard disk, for example), then
by default the computer moves the file to that location. If you drag the file to a location on
a different device (from a floppy to a hard disk, for instance), the computer automatically
copies the file, placing the new, identically named copy on the device.
UNIX lacks this subtlety. Instead, UNIX lets you choose which of the two operations you’d
like to perform. The two commands are typically succinct UNIX mnemonics: mv to move
files, and cp to copy files. The mv command also serves the dual duty of enabling you to rename
files.
JUST A MINUTE
Pronounce cp as “sea pea.” When you talk about copying a file, however, say “copy.” Similarly, pronounce mv as “em vee,” but when you
speak of moving a file, say “move.”
I find myself using cp more than mv because it offers a slightly safer way to organize files: if
I get confused and rename it such that it steps on another file (you’ll see what I mean in a
moment), I still have original copies of all the files.
6
117
1. The format of a cp command is to specify first the name of the file you want to
copy and then the new filename. Both names must be either relative filenames (that
is, without a leading slash or other indication of the directory) or absolute
filenames. Start out by making a copy of your .login file, naming the new copy
login.copy :
% cp .login login.copy
% ls -ld .login login.copy
-rw------- 1 taylor
1858 Oct 12 21:20 .login
-rw------- 1 taylor
1858 Nov 5 12:08 login.copy
You can see that the new file is identical in size and permissions but that it has a
more recent creation date, which certainly makes sense.
2. What happens if you try to copy a directory?
% cp . newdir
cp: .: Is a directory (not copied).
Generally, UNIX won’t permit you to use the cp command to copy directories.
JUST A MINUTE
I found that this command worked—sort of—on one machine I have used.
The system’s response to the cp command indicated that something
peculiar was happening with the following message:
cp: .: Is a directory (copying as plain file)
But, the system also created newdir as a regular, executable file. You may
find that your system reacts in this manner, but you probably do not have
any use for it.
3. The cp command is quite powerful, and it can copy many files at once if you
specify a directory as the destination rather than specifying a new filename.
Further, if you specify a directory destination, the program automatically will create
new files and assign them the same names as the original files.
First, you need to create a second file to work with:
% cp .cshrc cshrc.copy
Now try it yourself. Here is what I did:
% cp login.copy cshrc.copy NEWDIR
% ls -l NEWDIR
total 4
-rw------- 1 taylor
1178 Nov
-rw------- 1 taylor
1858 Nov
5 12:18 cshrc.copy
5 12:18 login.copy
5
6
118
Hour 6
You can use the cp command to copy an original file as a new file or to a specific
directory (the format being cp original-file new-file-or-directory), and you can
copy a bunch of files to a directory (cp list-of-files new-directory). Experiment with
creating new directories using mkdir and copying the files into the new locations. Use ls to
confirm that the originals aren’t removed as you go along.
Task 6.3: Moving Files to New Locations Using mv
Whereas cp leaves the original file intact, making a sort of electronic equivalent of
a photocopy of a paper I may pick up at my desk, mv functions like a more traditional
desk: papers are moved from one location to another. Rather than creating multiple copies
of the files you’re copying, mv physically relocates them from the old directory to the new.
1. You use mv almost the same way that you use cp:
% ls -l login.copy
-rw------- 1 taylor
1858 Nov
% mv login.copy new.login
% ls -l login.copy new.login
login.copy not found
-rw------- 1 taylor
1858 Nov
5 12:08 login.copy
5 12:08 new.login
2. Also, you move a group of files together using mv almost the same way you do it
using cp:
% cd NEWDIR
% ls
cshrc.copy login.copy
% mv cshrc.copy login.copy ..
% ls -l
total 0
% ls ..
Archives/
OWL/
InfoWorld/
PubAccessLists.Z
LISTS
bin/
Mail/
cshrc.copy
NEWDIR/
educ
News/
login.copy
new.login
rumors.26Oct.Z
rumors.5Nov.Z
src/
3. Because you can use mv to rename files or directories, you can relocate the new
directory NEWDIR. However, you cannot use mv to relocate the dot directory because
you’re inside it:
% mv . new.dot
mv: .: rename: Invalid argument
4. Both mv and cp can be dangerous. Carefully consider the following example before
trying either mv or cp on your own computer:
% ls -l login.copy cshrc.copy
-rw------- 1 taylor
1178 Nov 5 12:38 cshrc.copy
-rw------- 1 taylor
1858 Nov 5 12:37 login.copy
% cp cshrc.copy login.copy
% ls -l .login login.copy cshrc.copy
6
-rw-------rw-------
1 taylor
1 taylor
1178 Nov
1178 Nov
119
5 12:38 cshrc.copy
5 12:38 login.copy
Without bothering to warn me, UNIX copied the file cshrc.copy over the existing
file login.copy. Notice that after the cp operation occurred, both files had the same
size and modification dates.
The mv command will cause the same problem:
% ls -l cshrc.copy login.copy
-rw------- 1 taylor
1178 Nov
-rw------- 1 taylor
1858 Nov
% mv cshrc.copy login.copy
% ls -l cshrc.copy login.copy
cshrc.copy not found
-rw------- 1 taylor
1178 Nov
JUST A MINUTE
5 12:42 cshrc.copy
5 12:42 login.copy
5 12:42 login.copy
The good news is that you can set up UNIX so it won’t overwrite files. The
bad news is that for some reason many systems don’t default to this
behavior. If your system is configured reasonably, when you try either of
the two preceding dangerous examples, the system’s response is remove
login.copy? You can either press the Y key to replace the old file or press
Enter to change your mind. If your system cannot be set up to respond this
way, you can use the -i flag to both cp and mv to avoid this problem.
Later, you learn how to permanently fix this problem with a shell alias.
Together, mv and cp are the dynamic duo of UNIX file organization. These
commands enable you to put the information you want where you want it, leaving
duplicates behind if desired.
5
Task 6.4: Renaming Files with mv
Both the DOS and Macintosh systems have easy ways to rename files. In DOS, you
can use RENAME to accomplish the task. On the Mac, you can select the name under
the file icon and enter a new filename.
UNIX has neither option. To rename files, you use the mv command, which, in essence,
moves the old name to the new name. It’s a bit confusing, but it works.
1. Rename the file cshrc.copy with your own first name. Here’s an example:
% ls -l cshrc.copy
-rw------- 1 taylor
% mv cshrc.copy dave
% ls -l dave
-rw------- 1 taylor
1178 Nov
5 13:00 cshrc.copy
1178 Nov
5 13:00 dave
6
Hour 6
120
2. Rename a directory, too:
% ls -ld NEWDIR
drwxrwx--- 2 taylor
512 Nov
% mv NEWDIR New.Sample.Directory
% ls -ld New.Sample.Directory
drwxrwx--- 2 taylor
512 Nov
5 12:32 NEWDIR/
5 12:32 New.Sample.Directory/
3. Be careful! Just as moving files with cp and mv can carelessly overwrite existing files,
renaming files using mv can overwrite existing files:
% mv dave login.copy
If you try to use mv to rename a directory with a name that already has been
assigned to a file, the command fails:
% mv New.Sample.Directory dave
mv: New.Sample.Directory: rename: Not a directory
The reverse situation works fine because the file is moved into the directory as
expected. It’s the subtlety of using the mv command to rename files.
4. If you assign a new directory a name that belongs to an existing directory, some
versions of mv will happily overwrite the existing directory and name the new one as
requested:
% mkdir testdir
% mv New.Sample.Directory testdir
Being able to rename files is another important part of building a useful UNIX virtual
file cabinet for yourself. There are some major dangers involved, however, so tread
carefully and always use ls in conjunction with cp and mv to ensure that in the process you
don’t overwrite or replace an existing file.
Task 6.5: Removing Directories with rmdir
Now that you can create directories with the mkdir command, it’s time to learn how
to remove directories using the rmdir command.
1. With rmdir, you can remove any directory for which you have appropriate permissions:
% mkdir test
% ls -l test
total 0
% rmdir test
Note that the output of ls shows there are no files in the test directory.
6
121
2. The rmdir command removes only directories that are empty:
% mkdir test
% touch test/sample.file
% ls -l test
total 0
-rw-rw---- 1 taylor
0 Nov
% rmdir test
rmdir: test: Directory not empty
5 14:00 sample.file
To remove a directory, you must first remove all files therein using the rm command. In this example, test still has files in it.
3. Permissions are important, too. Consider what happens when I try to remove a
directory that I don’t have permission to touch:
% rmdir /tmp
rmdir: /tmp: Permission denied
% ls -l /tmp
drwxrwxrwt 81 root
15872 Nov
5 14:07 /tmp/
The permissions of the parent directory, rather than the directory you’re trying to
remove, are the important consideration.
There’s no way to restore a directory you’ve removed, so be careful and think through
what you’re doing. The good news is that, because with rmdir you can’t remove a
directory having anything in it (a second reason the attempt in the preceding example to
remove /tmp would have failed), you’re reasonably safe from major gaffes. You are not safe,
however, with the next command, rm, because it will remove anything.
Task 6.6: Removing Files Using rm
The rm command is the most dangerous command in UNIX. Lacking any sort of
archival or restoration feature, the rm command removes files permanently. It’s like
throwing a document into a shredder instead of into a dustbin.
1. Removing a file using rm is easy. Here’s an example:
% ls -l login.copy
-rw------- 1 taylor
% rm login.copy
% ls -l login.copy
login.copy not found
1178 Nov
5 13:00 login.copy
If you decide that you removed the wrong file and actually wanted to keep the
login.copy file, it’s too late. You’re out of luck.
5
6
Hour 6
122
2. You can remove more than one file at a time by specifying each of the files to the
rm command:
% ls
Archives/
PubAccessLists.Z
InfoWorld/
bin/
LISTS
cshrc.copy
Mail/
educ
News/
login.copy
OWL/
% rm cshrc.copy login.copy new.login
% ls
Archives/
OWL/
InfoWorld/
PubAccessLists.Z
LISTS
bin/
Mail/
educ
News/
new.login
rumors.26Oct.Z
rumors.5Nov.Z
src/
test/
testdir/
rumors.26Oct.Z
rumors.5Nov.Z
src/
test/
testdir/
3. Fortunately, rm does have a command flag that to some degree helps avoid accidental file removal. When you use the -i flag to rm (the i stands for interactive in this
case), the system will ask you if you’re sure you want to remove the file:
% touch testme
% rm -i testme
rm: remove testme? n
% ls testme
testme
% rm -i testme
rm: remove testme? y
% ls testme
testme not found
Note that n is no and y is yes. Delete the file.
4. Another flag that is often useful for rm, but is very dangerous, is the -r flag for
recursive deletion of files (a recursive command repeatedly invokes itself). When the
-r flag to rm is used, UNIX will remove any specified directory along with all its
contents:
% ls -ld test
drwxrwxrwx 3
total 1
-rw-rw---- 1
drwxrwx--- 2
; ls -lR test
taylor
512 Nov
5 15:32 test/
taylor
taylor
5 15:32 alpha
5 15:32 test2/
test/test2:
total 0
-rw-rw---- 1 taylor
% rm -r test
% ls -ld test
test not found
0 Nov
512 Nov
0 Nov
5 15:32 file1
Without any warning or indication that it was going to do something so drastic,
entering rm -r test caused not just the test directory, but all files and directories
inside it as well, to be removed.
6
JUST A MINUTE
123
This latest example demonstrates that you can give several commands in a
single UNIX command line. To do this, separate the commands with a
semicolon. Instead of giving the commands ls -ld test and ls -lR test
on separate lines, I opted for the more efficient ls -ld test; ls -lR test,
which uses both commands at once.
The UNIX equivalent of the paper shredder, the rm command allows easy removal
of files. With the -r flag, you can even clean out an entire directory. Nothing can be
retrieved after the fact, however, so use great caution.
Task 6.7: Minimizing the Danger of the rm
Command
At this point, you might be wondering why I am making such a big deal of the rm
command and the fact that it does what it is advertised to do—that is, remove files.
The answer is that learning a bit of paranoia now can save you immense grief in the future.
It can prevent you from destroying a file full of information you really needed to save.
For DOS, there are commercial programs (Norton Utilities, for instance) that can retrieve
accidentally removed files. The trash can on the Macintosh can be clicked open and the files
retrieved with ease. If the trash can is emptied after a file is accidentally discarded, a program
such as Symantec Utilities for the Macintosh can be used to restore files.
UNIX just doesn’t have that capability, though, and files that are removed are gone forever.
The only exception is if you work on a UNIX system that has an automatic, reliable backup
schedule. In such a case, you might be able to retrieve from a storage tape an older version
of your file (maybe).
5
That said, there are a few things you can do to lessen the danger of using rm and yet give
yourself the ability to remove unwanted files.
6
1. You can use a shorthand, a shell alias, to attach the -i flag automatically to each use
of rm. To do this, you need to ascertain what type of login shell you’re running,
which you can do most easily by using the following command. (Don’t worry
about what it all does right now. You learn about the grep command a few hours
from now.)
% grep taylor /etc/passwd
taylor:?:19989:1412:Dave Taylor/users/taylor:/bin/csh
Hour 6
124
The last word on the line is what’s important. The /etc/passwd file is one of the
database files UNIX uses to track accounts. Each line in the file is called a password
entry or password file entry. On my password entry, you can see that the login shell
specified is /bin/csh. If you try this and you don’t have an identical entry, you
should have /bin/sh or /bin/ksh.
2. If your entry is /bin/csh, enter exactly what is shown here:
% echo “alias rm /bin/rm -i” >> ~/.cshrc
% source ~/.cshrc
Now rm includes the -i flag each time it’s used:
% touch testme
% rm testme
rm: remove testme? n
3. If your entry is /bin/ksh, enter exactly what is shown here, paying particular
attention to the two different quotation mark characters used in the example:
$ echo ‘alias rm=”/bin/rm -i”’ >> ~/.profile
$ . ~/.profile
Now rm includes the -i flag each time it’s used.
CAUTION
One thing to pay special attention to is the difference between the single
quote (‘), the double quote (“), and the backquote (`). UNIX interprets
each differently, although single and double quotes are often interchangeable. The backquotes, also known as grave accents, are more unusual
and delineate commands within other commands.
4. If your entry is /bin/sh, you cannot program your system to include the -i flag
each time rm is used. The Bourne shell, as sh is known, is the original command
shell of UNIX. The Bourne shell lacks an alias feature, a feature that both the Korn
shell (ksh) and the C shell (csh) include. As a result, I recommend that you change
your login shell to one of these alternatives, if available.
To see what’s available, look in the /bin directory on your machine for the specific
shells:
% ls -l /bin/sh /bin/ksh /bin/csh
-rwxr-xr-x 1 root
102400 Apr 8 1991 /bin/csh*
-rwxr-xr-x 1 root
139264 Jul 26 14:35 /bin/ksh*
-rwxr-xr-x 1 root
28672 Oct 10 1991 /bin/sh*
Most of the examples in this book focus on the C Shell because I think it’s the
easiest of the three shells to use. To change your login shell to csh, you can use the
chsh—change login shell—command:
% chsh
Changing login shell for taylor.
6
125
Old shell: /bin/sh
New shell: /bin/csh
Now you can go back to instruction 2 and set up a C shell alias. This will help you
avoid mischief with the rm command.
The best way to avoid trouble with any of these commands is to learn to be just a bit
paranoid about them. Before you remove a file, make sure it’s the one you want.
Before you remove a directory, make doubly sure that it doesn’t contain any files you might
want. Before you rename a file or directory, double-check to see if renaming it is going to
cause any trouble.
Take your time with the commands you learned in this hour, and you should be fine. Even
in the worst case, you might have the safety net of a system backup performed by a system
administrator, but don’t rely on it.
Summary
You now have completed six hours of UNIX instruction, and you are armed with enough
commands to cause trouble and make UNIX do what you want it to do. In this hour, you
learned the differences between cp and mv for moving files and how to use mv to rename both
files and directories. You also learned how to create directories with the mkdir command and
how to remove them with the rmdir command. And you learned about the rm command for
removing files and directories, and how to avoid getting into too much trouble with it.
Finally, if you were really paying attention, you learned how to identify which login shell
you’re using (csh, ksh, or sh) and how to change from one to another using the chsh
command.
5
Workshop
in the next hour.
Key Terms
password entry For each account on the UNIX system, there is an entry in the account
database known as the password file. This also contains an encrypted copy of the account
password. This set of information for an individual account is known as the password entry.
recursive command A command that repeatedly invokes itself.
6
Hour 6
126
shell alias Most UNIX shells have a convenient way for you to create abbreviations for
commonly used commands or series of commands, known as shell aliases. For example, if
I always found myself typing ls -CF, an alias can let me type just ls and have the shell
automatically add the -CF flags each time.
Questions
1. What are the differences between cp and mv?
2. If you were installing a program from a floppy disk onto a hard disk, would you
use cp or mv?
3. If you know DOS, this question is for you. Although DOS has a RENAME command,
it doesn’t have both COPY and MOVE. Which of these two do you think DOS
includes? Why?
4. Try using mkdir to create a directory. What happens and why?
5. You’ve noticed that both rmdir and rm -r can be used to remove directories.
Which is safer to use?
6. The rm command has another flag that wasn’t discussed in this hour. The -f flag
forces removal of files regardless of permission (assuming you’re the owner, that is).
In combination with the -r flag, this can be amazingly destructive. Why?
The seventh hour introduces the useful file command, which indicates the contents of any
file in the UNIX file system. With file, you will explore various directories in the UNIX file
system to see what it reveals about different system and personal files. Then, when you’ve
found some files worth reading, you will learn about cat, more, and pg, which are different
ways of looking at the contents of a file.
6
Looking into Files
Hour
127
7
Looking into Files
By this point, you’ve learned a considerable number of UNIX commands and
a lot about the operating and file systems. This hour focuses on UNIX tools to
help you ascertain what type of files you’ve been seeing in all the different
directories. It then introduces five powerful tools for examining the content
of files.
Goals for This Hour
■
■
■
■
■
■
Use file to identify file types
Explore UNIX directories with file
Peek at the first few lines with head
View the last few lines with tail
View the contents of files with cat
View larger files with more
This hour begins with a tool to help ensure that the files you’re about to view
are intended for human perusal and then explores many of the commands
available to view the contents of the file in various ways.
7
Hour 7
128
Task 7.1: Using file to Identify File Types
One of the most undervalued commands in UNIX is file, which is often neglected
and collecting dust in some corner of the system. The file command is a program
that can easily offer you a good hint as to the contents of a file by looking at the first few lines.
Unfortunately, there is a problem with the file command: It isn’t 100 percent accurate. The
program relies on a combination of the permissions of a file, the filename, and an analysis of
the first few lines of the text. If you had a text file that started out looking like a C program
or that had execute permission enabled, file might well identify it as an executable program
rather than an English text file.
JUST A MINUTE
You can determine how accurate your version of file is by checking the
size of its database of file types. You can do this with the UNIX command
wc -l /etc/magic . The number of entries in the database should be
around 100. If you have many less than this number, you’re probably
going to have trouble. If you have considerably more, you might have a
very accurate version of file at your fingertips! Remember, however,
even if it’s relatively small, file can still offer invaluable suggestions
regarding file content anyway.
1. Start by logging in to your account and using the ls command to find a file or two
to check.
% ls -F
Archives/
InfoWorld/
LISTS
Mail/
News/
OWL/
PubAccessLists.Z
bin/
educ
rumors.26Oct.Z
rumors.5Nov.Z
src/
temp/
Next, simply enter the file command, listing each of the files you’d like the
program to analyze:
% file LISTS educ rumors.26Oct.Z src
LISTS: ascii text
educ:
ascii text
rumors.26Oct.Z: block compressed 16 bit code data
src:
directory
From this example, you can see that file correctly identifies src as a directory,
offers considerable information on the compressed file rumors.26Oct.Z, and tags
both LISTS and educ as plain ASCII text files.
7
Looking into Files
JUST A MINUTE
129
ASCII is the American Standard Code for Information Interchange and
means that the file contains the letters of the English alphabet, punctuation,
and numbers, but not much else. There are no multiple typefaces, italics,
or underlined passages, and there are no graphics. It’s the lowest
common denominator of text in UNIX.
2. Now try using the asterisk (*), a UNIX wildcard (explained in Hour 9, “Wildcards
and Regular Expressions”), to have the program analyze all files in your home
directory:
% file *
Global.Software:
English text
Interactive.Unix:
mail folder
Mail:
directory
News:
directory
Src:
directory
bin:
directory
history.usenet.Z:
compressed data block compressed 16 bits
The asterisk (*) is a special character in UNIX. Used by itself, it tells the system to
replace it with the names of all the files in the current directory.
This time you can begin to see how file can help differentiate files. Using this
command, I am now reminded that the file Global.Software is English text, but
Interactive.Unix is actually an old electronic mail message (file can’t differentiate
between a single mail message and a multiple-message folder, so it always errs on
the side of saying that the file is a mail folder).
3. Mail folders are actually problematical for the file command. On one of the
systems I use, the file command doesn’t know what mail messages are, so asking it
to analyze mail folders results in a demonstration of how accuracy is related to the
size of the file database.
On a Sun system, I asked file to analyze two mail folders, with the following
results:
% file Mail/mailbox Mail/sent
Mail/mailbox:
mail folder
Mail/sent: mail folder
Those same two files on a Berkeley UNIX system, however, have very different
results when analyzed:
% file Mail/mailbox Mail/sent Mail/netnews
Mail/mailbox:
ascii text
Mail/sent:
shell commands
Mail/netnews:
English text
Not only does the Berkeley version of UNIX not identify the files correctly, it
doesn’t even misidentify them consistently.
7
Hour 7
130
4. Another example of the file command’s limitations is how it interacts with file
permissions. Use cp to create a new file and work through this example to see how
your file command interprets the various changes.
% cp .cshrc test
% file test
test: shell commands
% chmod +x test
% file test
test: shell script
Adding execute permission to this file caused file to identify it as a shell script
rather than shell commands.
Don’t misinterpret the results of these examples as proof that the file command is
useless and that you shouldn’t use it. Quite the opposite is true. UNIX has neither
a specific file-naming convention (DOS has its three-letter filename suffixes) nor indication
of file ownership by icon (Macintosh does this with creator information added by each
program). As a result, it’s vital that you have a tool for helping ascertain file types without
actually opening the file.
Why not just look at the contents? The best way to figure out the answer to this question is
to display accidentally the contents of an executable file on the screen. You’ll see it’s quite a
mess, loaded with special control characters that can be best described as making your screen
go berserk.
Task 7.2: Exploring UNIX Directories with file
Now that you know how to work with the file command, it’s time to wander
through the UNIX file system, learning more about types of files that tend to be
found in specific directories. Your system might vary slightly—it’ll certainly have more files
in some directories than what I’m showing here in the examples, but you’ll quickly see that
file can offer some valuable insight into the contents of files.
1. First things first. Take a look at the files found in the very top level of the file
system, in / (root):
% cd /
% ls -CF
-No _rm_ star boot
flags/
rhf@
userb/
OLD/
core
gendynix
stand/
userc/
archive/
dev/
lib@
sys@
userd/
ats/
diag@
lost+found/ tftpboot@
usere/
backup/
dynix
mnt/
tmp/
users/
bin@
etc/
net/
usera/
usr/
% file boot core gendynix tftpboot
boot:
SYMMETRY i386 stand alone executable version 1
7
Looking into Files
131
core:
core from getty
gendynix:
SYMMETRY i386 stand alone executable not
➥stripped version 1
tftpboot:
symbolic link to /usr/tftpboot
This example is from a Sequent computer running DYNIX, the Sequents’ version
of UNIX, based on Berkeley 4.3 BSD with some AT&T System V extensions. It’s
the same machine that has such problems identifying mail folders.
Executable binaries are explained in detail by the file command on this computer:
boot is listed as SYMMETRY i386 stand alone executable version 1. The specifics
aren’t vital to understand: The most important word to see in this output is
executable , indicating that the file is the result of compiling a program. The
format is SYMMETRY i386, version 1, and the file requires no libraries or other files
to execute—it’s stand-alone.
For gendynix, the format is similar, but one snippet of information is added that
isn’t indicated for boot: The executable file hasn’t been stripped.
JUST A MINUTE
Stripping a file doesn’t mean that you peel its clothes off, but rather that a
variety of information included in most executables to help identify and
isolate problems has been removed to save space.
When a program dies unexpectedly in UNIX, the operating system tries to leave a
snapshot of the memory that the program was using, to aid in debugging. Wading
through these core files can be quite difficult—it’s usually reserved for a few experts
at each site, but there is still some useful information inside. The best, and simplest, way to check it is with the file command. You can see in the preceding
listing that file recognized the file core as a crashed program memory image and
further extracted the name of the program that originally failed, getty, causing the
program to fail. When this failure happens, UNIX creates an image of the program
in memory at the time of failure, which is called a core dump.
The fourth of the listings offers an easy way to understand symbolic links, indicated
in ls -CF output with the special suffix @, as shown in the preceding example with
tftpboot@. Using file , you can see that the file tftpboot in the root directory is
actually a symbolic link to a file with the same name elsewhere in the file system,
/usr/tftpboot .
2. There are differences in output formats on different machines. The following
example shows what the same command would generate on a Sun Microsystems
workstation, examining analogous files:
% file boot core kadb tmp
boot:
sparc executable
core:
core file from ‘popper’
7
Hour 7
132
kadb:
tmp:
sparc executable not stripped
symbolic link to /var/tmp
The Sun computer offers the same information but fewer specifics about executable
binaries. In this case, Sun workstations are built around SPARC chips (just like
PCs are built around Intel chips), so the executables are identified as sparc
executable .
3. Are you ready for another directory of weird files? It’s time to move into the /lib
directory to see what devices are present on your system and what type of files
they are.
Entering ls will demonstrate quickly that there are a lot of files in this directory!
The file command can tell you about any of them. On my Sun computer, I asked
for information on a few select files, many of which you might also have on yours:
% file lib.b lib300.a diffh sendmail
lib.b:
c program text
lib300.a:
archive random library
diffh:
sparc pure dynamically linked executable not stripped
sendmail:
sparc demand paged dynamically linked set-uid executable
The first file, lib.b, demonstrates that the file command works regardless of the
name of a file: Standard naming for C program files specifies that they end with the
characters .c, as in test.c. So, without file, you might never have suspected that
lib.b is a C program. The second file is an actual program library and is identified
here as an archive random library, meaning that it’s an archive and that the
information within can be accessed in random order (by appropriate programs).
The third file is an executable, demonstrating another way that file can indicate
programs on a Sun workstation. The sendmail program is an interesting program:
It’s an executable, but it has some new information that you haven’t seen before.
The set-uid indicates that the program is set up so that when anyone runs it,
sendmail runs as the user who owns the file, not the user who launched the
program. A quick ls can reveal a bit more about this:
% ls -l /lib/sendmail
-r-sr-x--x 1 root
155648 Sep 14 09:11 /lib/sendmail*
Notice here that the fourth character of the permissions string is an s rather than
the expected x for an executable. Also check the owner of the file in this listing.
Combined, the two mean that when anyone runs this program, sendmail actually
will set itself to a different user ID (root in this case) and have that set of access
permissions. Having sendmail run with root permissions is how you can send
electronic mail to someone else without fuss, but you can’t view his or her mailbox.
4. Consider now one more directory full of weird files before you start the next task.
This time, move into the /dev directory and see what’s inside. Again, it’s a directory with a lot of files, so don’t be surprised if the output scrolls off the screen!
7
Looking into Files
133
Try to identify a few files that are similar in name to the ones I examine here, and
see what file says about them:
% cd /dev
% file MAKEDEV audio spx sr0 tty
MAKEDEV:
executable shell script
audio:
character special (69/0)
spx:
sr0:
block special (18/0)
tty:
UNIX has two different types of devices, or peripherals, that can be attached: those
that expect information in chunks and those that are happier working on a byte-bybyte basis. The former are called block special devices and the latter character special
devices. You don’t have to worry about the differences, but notice that file can
differentiate between them: audio, spx, and tty are all character-special-device files,
whereas sr0 is a block-special-device file.
The pair of numbers in parentheses following the description of each file are
known as the major number and minor number of the file. The first indicates the
type of device, and the second indicates the physical location of the plug, wire,
card, or other hardware that is controlled by the specific peripheral.
The good news is that you don’t have to worry a bit about what files are in the /lib,
/etc, or any other directory other than your own home directory. There are
thousands of happy UNIX folk working busily away each day without ever realizing that these
other directories exist, let alone knowing what’s in them.
What’s important here is that you have learned that the file command is quite sophisticated
at identifying special UNIX system files of various types. It can be a very helpful tool when
you are looking around in the file system and even when you are just trying to remember
which files are which in your own directory.
Task 7.3: Peeking at the First Few Lines with head
Now that you have the tools needed to move about in the file system, to double check
where you are, and to identify the types of different files, it’s time to learn about some
of the many tools UNIX offers for viewing the contents of files. The first on the list is head,
a simple program for viewing the first ten lines of any file on the system.
The head program is more versatile than it sounds: you can use it to view up to the first few
hundred lines of a very long file, actually. To specify the number of lines you want to see, you
just need to indicate how many as a starting argument, prefixing the number of lines desired
with a dash.
7
Hour 7
134
JUST A MINUTE
This command, head, is the first of a number of UNIX commands that tend
to work with their own variant on the regular rules of starting arguments.
Instead of a typical UNIX command argument of -l33 to specify 33 lines,
head uses -33 to specify the same information.
1. Start by moving back into your home directory and viewing the first few lines of
your .cshrc file:
% cd
% head .cshrc
#
# Default user .cshrc file (/bin/csh initialization).
set host=limbo
set path=(. ~/bin /bin /usr/bin /usr/ucb /usr/local /etc
/usr/etc/usr/local/bin /usr/unsup/bin)
# Set up C shell environment:
alias
diff
‘/usr/bin/diff -c -w’
The contents of your own .cshrc file will doubtless be different, but notice that the
program lists only the first few lines of the file.
2. To specify a different number of lines, use the -n format (where n is the number of
lines). I’ll look at just the first four lines of the .login file:
% head -4 .login
#
# @(#) $Revision: 62.2 $
setenv TERM vt100
3. You also can easily check multiple files by specifying them to the program:
% head -3 .newsrc /etc/passwd
==> .newsrc <==
misc.forsale.computers.mac: 1-14536
utech.student-orgs! 1
general! 1-546
==> /etc/passwd <==
root:?:0:0: root,,,,:/:/bin/csh
news:?:6:11:USENET News,,,,:/usr/spool/news:/bin/ksh
ingres:*?:7:519:INGRES Manager,,,,:/usr/ingres:/bin/csh
7
4. More importantly, head, and other UNIX commands, can work also as part of a
pipeline, where the output of one program is the input of the next. The special
symbol for creating UNIX pipelines is the pipe (|) character. Pipes are read left to
Looking into Files
135
right, so you can easily have the output of who, for example, feed into head, offering
powerful new possibilities. Perhaps you want to see just the first five people logged
in to the computer right now. Try this:
% who | head -5
root
console
mccool
ttyaO
millekl2 ttyaP
paulwhit ttyaR
bobweir ttyaS
Broken pipe
Nov
Nov
Nov
Nov
Nov
9
10
10
10
10
07:31
14:25
14:58
14:50
14:49
Pipelines are one of the most powerful features of UNIX, and there are many
examples of how to use them to best effect throughout the remainder of this book.
5. Here is one last thing. Find an executable, /boot will do fine, and enter head -1 /
boot. Watch what happens. Or, if you’d like to preserve your sanity, take it from
me that the random junk thrown on your screen is plenty to cause your program to
get quite confused and possibly even quit or crash.
The point isn’t to have that happen to your screen, but rather to remind you that
using file to confirm file type for unfamiliar files can save you lots of grief and
frustration!
The simplest of programs for viewing the contents of a file, head, is easy to use,
efficient, and works as part of a pipeline, too. The remainder of this hour focuses on
other tools in UNIX that offer other ways to view the contents of text and ASCII files.
Task 7.4: Viewing the Last Few Lines with tail
The head program shows you the first 10 lines of the file you specify. What would
you expect tail to do, then? I hope you guessed the right answer: It shows the last
10 lines of a file. Like head, tail also understands the same format for specifying the number
of lines to view.
1. Start out viewing the last 12 lines of your .cshrc file:
% tail -12 .cshrc
set noclobber history=100 system=filec
umask 007
7
setprompt
endif
# special aliases:
alias info
alias ssinfo
ssinfo
‘echo “connecting...” ; rlogin oasis’
Hour 7
136
2. Next, the last four lines of the file LISTS in my home directory can be shown with
the following command line:
% tail -5 LISTS
College of Education
Arizona State University
Tempe, AZ 85287-2411
602-965-2692
Don’t get too hung up trying to figure out what’s inside my files: I’m not even sure
myself sometimes.
3. Here’s one to think about. You can use head to view the first n lines of a file and
tail to view the last n lines of a file. Can you figure out a way to combine the two
so you can see just the tenth, eleventh, and twelfth lines of a file?
% head
alias
alias
alias
-12 .cshrc | tail -3
diff
from
‘frm -n’
ll
‘ls -l’
It’s easy with UNIX command pipelines!
Combining the two commands head and tail can give you considerable power in
viewing specific slices of a file on the UNIX system. Try combining them in different
ways for different effects.
Task 7.5: Viewing the Contents of Files with cat
Both head and tail offer the capability to view a piece of a file, either the top or
bottom, but neither lets you see the entire file, regardless of length. For this job, the
cat program is the right choice.
JUST A MINUTE
The cat program got its name from its function in the early versions of
UNIX; its function was to concatenate (or join together) multiple files. It
isn’t, unfortunately, homage to feline pets or anything else so exotic!
The cat program also has a valuable secret capability, too: Through use of the -v flag, you
can use cat to display any file on the system, executable or otherwise, with all characters that
normally would not be printed (or would drive your screen bonkers) displayed in a special
format I call control-key notation. In control key notation, each character is represented as ^n,
where n is a specific printable letter or symbol. A character with the value of 0 (also referred
to as a null or null character) is displayed as ^@, a character with the value 1 is Â, a character
with the value 2 is ^B, and so on.
Another cat flag that can be useful for certain files is -s, which suppresses multiple blank lines
from a file. It isn’t immediately obvious how that could help, but there are some files that can
7
Looking into Files
137
have a screen full (or more) of blank lines. To avoid having to watch them all fly past, you
can use cat -s to chop ’em all down to a single blank line.
1. Move back to your home directory again, and use cat to display the complete
contents of your .cshrc file:
% cd
% cat .cshrc
#
/usr/etc/usr/local/bin /usr/unsup/bin )
alias
alias
alias
alias
alias
alias
diff
from
ll
ls
mail
mailq
‘frm -n’
‘ls -l’
‘/bin/ls -F’
Mail
‘/usr/lib/sendmail -bp’
alias
newaliases ‘echo you mean newalias...’
alias
alias
rd
rn
‘readmsg $ | page’
‘/usr/local/bin/rn -d$HOME -L -M -m -e -S -/’
# and some special stuff if we’re in an interactive shell
if ( $?prompt ) then
alias
alias
alias
cd
env
setprompt
# shell is interactive.
‘chdir \!* ; setprompt’
‘printenv’
‘set prompt=”$system ($cwd:t) \! : “‘
set noclobber history=100 system=limbo filec
umask 007
setprompt
endif
# special aliases:
alias info
alias ssinfo
ssinfo
Don’t be too concerned if the content of your .cshrc file (or mine) doesn’t make
any sense to you. You are slated to learn about the contents of this file within a few
hours, and, yes, it is complex.
7
138
Hour 7
You can see that cat is pretty simple to use. If you specify more than one filename
to the program, it lists them in the order you specify. You can even list the contents
of a file multiple times by specifying the same filename on the command line
multiple times.
2. The cat program also can be used as part of a pipeline. Compare the following
command with my earlier usage of head and tail:
% cat LISTS | tail -5
College of Education
Arizona State University
Tempe, AZ 85287-2411
602-965-2692
3. Now find an executable file, and try cat
glimpse of the contents therein:
-v
in combination with head to get a
% cat -v /bin/ls | head -1
M-k”^@^@^@M-^@^@^@^@^P^@^@M-45^@^@^@^@^@^@M-l^P^@^@^@^@^@^@
➥^@^@^@^@^@^@^@^@
^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@
➥^@^@^@^@^@^@^@^@^@
^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@^@
➥^@^@^@^@^@^@^@^@
^@^@^@^@^@^@^@^@^@^@^@^@Â^@^@^@$Header: crt0.c 1.4 87/04/23
➥$^@^@@(#)Copy
right (C) 1984 XXXXXXX Computer Systems, Inc. All rights reserved.
➥^@M-^KM-NM-^KMtM-^MF^DM-^KM-XM-^K^F@M-^M^DM-Ê^@^@^@^@M-^KM-SÂM-BM-ÎÛ^@M-^@^@
➥^@SM-^?6M-hw^T^@^@M^CM-D^HM-^?5^@M-^@^@^@SM-^?6M-h&^@^@^@M-^CM-D^LPM-h)[^@^@YM-tM-^PM➥^PM-^PM-k^BM-IM-CUM^KM-lM-kM-yM-^PM-^PM-^PM
-k^BM-IM-CUM-^KM-lM-kM-yM-^PM-^PM-^PUM-^KM-lM-^CM-l^XWVSM-^K
➥u^LM-^K]^HKM-^CM-F^DM-hM
-X^V^@^@M-ÊM-@u^FM-^?ÊM-lM-^L^@^@h^DM-^M^@^@M-hM-P7^@^
➥@YM-^KÊ^DM-^M^@^@-^@NMm^@M-ÎÊ^HM-^M^@^@M-^KÊ^DM-^M^@^@Ê^P^N^@^@M-ÎÊ^LM-^
➥M^@^@M-^CÊ^DM-^M^@^@<MGÊ^PM-^M^@^@P^@^@^@jÂM-hM-%[^@^@YM-ÊM-@tNh^TM-^M^@^@h
➥M-HM-^J^@^@M-h12^@^@M-^C
M-D^HM-GÊM-pM-^L^@^@Â^@^@^@M-8Â^@^@^@M-ÎÊM-hM-^L^@^
➥@M-ht^S^@^@M-^MEM-nPjÂM
-h]^V^@^@M-^CM-D^HÔM-?EM-r%^@^L^@^@=^@Broken pipe
This is complex and confusing, indeed! What’s worse, this isn’t the entire first line
of the executable. You can see that, because this block of data ends with Broken
pipe, which indicates that a lot more was being fed to head than it could process,
due to the constraint of having only the first line listed—a line that head defines as
no more than 512 characters long.
The cat command is useful for viewing files and is quite easy to use, too. The problem
with it is that if the file you choose to view has more than the number of lines on your
screen, the file will just fly past you without your having any way to slow it down. That’s where
7
Looking into Files
139
the next two commands come in handy: more for stepping through files, and page for paging
through files. Both solve this problem, albeit in slightly different ways.
Task 7.6: Viewing Larger Files with more
You can now wander about the file system, find files that might be of interest, check
their types with file, and even view them with the cat command, but what if they’re
longer than the size of your screen? That’s the job of the more program, a program that knows
how big your screen is and displays the information page by page.
There are three primary flags in more:
-s
-d
-c
Suppresses multiple blank lines, just like the -s flag to cat
Forces more to display friendlier prompts at the bottom of each page
Causes the program to clear the screen before displaying each screen full
of text
The program also allows you to start at a specific line in the file by using the curious +n
notation, where n is a specific number. Finally, you can start also at the first occurrence of
a specific pattern by specifying that pattern to the program in a format similar to +/pattern
(patterns are defined in Hour 9).
1. View the .cshrc file using more:
% more ~/.cshrc
#
set host=limbo
/usr/etc /usr/local/bin /usr/unsup/bin)
alias
alias
alias
alias
alias
alias
diff
from
ll
ls
mail
mailq
alias
alias rd
--More--(51%)
‘frm -n’
‘ls -l’
‘/bin/ls -F’
Mail
7
Hour 7
140
Unlike previous examples where the program runs until completed, leaving you
back on the command line, more is the first interactive program you’ve encountered.
When you see the --More--(51%) prompt, the cursor sits at the end of that line,
waiting for you to tell it what to do. The more program lets you know how far into
the file you’ve viewed, too; in the example, you’ve seen about half of the file (51
percent).
At this point, there is quite a variety of commands available. Press the spacebar to
see the next screen of information, until you have seen the entire file.
2. Try starting up the program with the twelfth line of the file:
% more +12 ~/.cshrc
alias mailq
alias
alias
alias
rd
rn
alias
alias
alias
cd
env
setprompt
‘printenv’
set noclobber history=100 filec
umask 007
setprompt
endif
--More--(82%)
3. You can see that about halfway through the .cshrc file there is a line that contains
the word newaliases. I can start up more so that the line with this pattern is
displayed on the top of the first screenful.
% more +/newaliases ~/.cshrc
...skipping
alias mailq
alias
alias
alias
rd
rn
7
Looking into Files
alias
alias
alias
cd
env
setprompt
141
‘printenv’
set noclobber history=100 filec
umask 007
setprompt
endif
# special aliases:
alias info
--More--(86%)
ssinfo
Actually, notice that the line containing the pattern newaliases shows up as the
third line of the first screen, not the first line. That’s so you have a bit of context to
the matched line, but it can take some getting used to. Also note that more tells
us—with the message ...skipping as the very first line—that it’s skipping some
lines to find the pattern.
4. The range of commands available at the --More-- prompt is quite extensive, as
listed in Table 7.1. The sidebar following the table explains what the conventions
used in the table mean and how to enter the following commands.
Table 7.1. Commands available within the more program.
Command
Function
[Space]
n[Return]
Press the spacebar to display the next screenful of text.
Display the next n lines (the default is the next line only of text).
Display a list of commands available in the more program.
Scroll down half a page.
Quit the more program.
Skip forward n lines (default is 1).
Skip forward n screenfuls (default is 1).
Skip backward a screenful of text.
Display the current line number.
Search for an occurrence of a pattern.
Search for the next occurrence of the current pattern.
Start the vi editor at the current line.
(That’s a lowercase L.) Redraw the screen.
Display the current filename and line number.
h
d
q
ns
nf
b
or Control-b
=
/pattern
n
v
Control-l
:f
7
Hour 7
142
Entering Commands in the more Program
In this table and in the following text, the word space enclosed in brackets [Space]
refers to pressing the spacebar as a command. Likewise, [Return] means you should
press the Return key as part of the command.
A hyphen in a command—for example Ctrl-b—means that you should hold down
the first indicated key while you press the second key. The lowercase-letter commands in the table indicate that you should press the corresponding key, the A key
for the a command, for example.
Two characters together, but without a hyphen (:f), mean that you should press
the appropriate keys in sequence as you would when typing text.
Finally, entries that have an n before the command mean that you can prefix the
command with a number, which will let it use that value to modify its action. For
example, 3[Return] displays the next three lines of the file and 250s skips the next
250 lines. Typically, pressing Return after typing a command within more is not
necessary.
Try some commands on a file of your own. A good file that will have enough lines
to make this interesting is /etc/passwd:
% more /etc/passwd
root:?:0:0: root:/:/bin/csh
news:?:6:11:USENET News:/usr/spool/news:/bin/ksh
ingres:*?:7:519:INGRES Manager:/usr/ingres:/bin/csh
usrlimit:?:8:800:(1000 user system):/mnt:/bin/false
vanilla:*?:20:805:Vanilla Account:/mnt:/bin/sh
charon:*?:21:807:The Ferryman:/users/tomb:
actmaint:?:23:809: Maintenance:/usr/adm/actmaint:/bin/ksh
pop:*?:26:819:,,,,:/usr/spool/pop:/bin/csh
lp:*?:70:10:System V Lp Admin:/usr/spool/lp:
trouble:*?:97:501:Trouble Report Facility:/usr/trouble:/usr/msh
postmaster:?:98:504:Mail:/usr/local/adm:/bin/csh
aab:?:513:1233:Robert Townsend:/users/aab:/bin/ksh
billing:?:516:1233:Accounting:/users/billing:/bin/csh
aai:?:520:1233:Pete Cheeseman:/users/aai:/bin/csh
--More--(1%) 60s
...skipping 60 lines
7
cq:?:843:1233:Rob Tillot:/users/cq:/usr/local/bin/tcsh
robb:?:969:1233:Robb:/users/robb:/usr/local/lib/msh
aok:?:970:1233:B Jacobs:/users/aok:/usr/local/lib/msh
went:?:1040:1233:David Math:/users/went:/bin/csh
aru:?:1076:1233:Raffie:/users/aru:/bin/ksh
varney:?:1094:1233:/users/varney:/bin/csh
brandt:?:1096:1233:Eric Brand:/users/brand:/usr/local/bin/tcsh
ask:?:1098:1233:/users/ask:/bin/csh
asn:?:1101:1233:Ketter Wesley:/users/asn:/usr/local/lib/msh
--More--(2%)
Looking into Files
143
This example isn’t exactly what you’ll see on your screen because each time you
type a command to more, it erases its own prompt and replaces the prompt with the
appropriate line of the file. Try pressing [Return] to move down one line, and
you’ll see what I mean.
Quit more in the middle of viewing this file by typing q.
The more program is one of the best general-purpose programs in UNIX, offering an
easy and powerful tool for perusing files. The biggest limitation, however, is that you
can’t go backward in the file: If you realize that what you are looking for is on the previous
page, you have to quit and start the program again.
Summary
Now that you can add this set of commands to your retinue of UNIX expertise, you are most
certainly ready to wander about your own computer system, understanding what files are
what, where they are, and how to peer inside. You learned about file to ascertain type, head
and tail for seeing snippets of files, and cat and more to help easily view files of any size on
your screen.
Workshop
in the next hour.
Key Terms
block special device A device driver that controls block-oriented peripherals. A hard disk,
for example, is a peripheral that works by reading and writing blocks of information (as
distinguished from a character special device). See also character special device.
character special device A device driver that controls a character-oriented peripheral.
Your keyboard and display are both character-oriented devices, sending and displaying
information on a character-by-character basis. See also block special device.
control-key notation A notational convention in UNIX that denotes the use of a control
key. There are three common conventions: Ctrl-C, ^c and C-C all denote the Control-c
character, produced by pressing the Control key (labeled Control or Ctrl on your keyboard)
and, while holding it down, pressing the c key.
core dump The image of a command when it executed improperly.
interactive program An interactive UNIX application is one that expects the user to enter
information and then responds as appropriate. The ls command is not interactive, but the
more program, which displays text a screenful at a time, is interactive.
7
Hour 7
144
major number For device drivers, the major number identifies the specific type of device
in use to the operating system. This is more easily remembered as the device ID number.
minor number Once the device driver is identified to the operating system by its major
number, the address of the device in the computer itself (that is, which card slot a peripheral
card is plugged into) is indicated by its minor number.
null character Each character in UNIX has a specific value, and any character with a
numeric value of zero is known as a null or null character.
pipeline
A series of UNIX commands chained by |, the pipe character.
Questions
1. Many people who use UNIX systems tend to stick with file-naming conventions.
Indeed, UNIX has many of its own, including .c for C source files, .Z for compressed files, and a single dot prefix for dot files. Yet file ignores filenames (test it
yourself). Why?
2. Use more to check some of the possible file types that can be recognized with the
file command by peeking in the configuration file /etc/magic .
3. Do you remember the television game show “Name that Tune?” If so, you’ll recall
how contestants had to identify a popular song by hearing just the first few notes.
The file command is similar; the program must guess at the type of the file by
checking only the first few characters. Do you think it would be more accurate by
checking more of the file, or less accurate? (Think about this one.)
4. How did the cat command get its name? Do you find that a helpful mnemonic?
5. Here’s an oddity: What will this command do?
cat LISTS | more
6. If you were looking at an absolutely huge file and you were pretty sure that what
you wanted was near the bottom, what command would you use, and why?
7. What if the information is near the top?
There are lots of special characters in UNIX, as you have doubtless learned by accidentally
typing a slash, asterisk, question mark, quote, or just about any other punctuation character.
What may surprise you is that they all have different, specific meanings. The next hour
explains considerably more about how pipelines work and how programs are used as filters.
Among the new commands you will learn are sort, wc, nl, uniq, and spell. You also will learn
a new, immensely helpful flag to cat that makes cat produce line numbers.
7
Filters and Piping
145
8
Hour
8
Filters and Piping
If you’ve ever learned a foreign language, you know that the most common
approach is to start by building your vocabulary (almost always including the
names of the months, for some reason), and then you learn about sentence
construction rules. The UNIX command line is a lot like a language. Now
you’ve learned a lot of UNIX words, so it’s time to learn how to put them
together as sentences using file redirection, filters, and pipes.
Commands to be added to your vocabulary this hour include wc, sort, nl, and
uniq. You also learn about the -n flag to the cat command, which forces cat to
add line numbers, and how you can use that to help find information within files.
Goals for This Hour
■
■
■
■
■
■
The secrets of file redirection
How to count words and lines using wc
How to remove extraneous lines using uniq
How to sort information in a file using sort
How to add line numbers to files with cat -n and nl
Cool nl tricks and capabilities
5
Hour 8
146
This hour begins by focusing on one aspect of constructing powerful custom commands in
UNIX by using file redirection. The introduction of some filters, programs that are intended
to be used as part of command pipes, follow. Next you learn another aspect of creating your
own UNIX commands using pipelines.
Task 8.1: The Secrets of File Redirection
So far, all the commands you’ve learned while teaching yourself UNIX have required
you to enter information at the command line, and all have produced output on the
screen. But, as Gershwin wrote in Porgy and Bess, “it ain’t necessarily so.” In fact, one of the
most powerful features of UNIX is that the input can come from a file as easily as it can come
from the keyboard, and the output can be saved to a file as easily as it can be displayed on your
screen.
The secret is file redirection, the special commands in UNIX that instruct the computer to
read from a file, write to a file, or even append information to an existing file. Each of these
acts can be accomplished by placing a file-redirection command in a regular command
line: < redirects input, > redirects output, and >> redirects output and appends the
information to the existing file. A mnemonic for remembering which is which is to remember
that, just as in English, UNIX works from left to right, so a character that points to the left
(<) changes the input, whereas a character that points right (>) changes the output.
1. Log in to your account and create an empty file using the touch command:
% touch testme
2. First, use this empty file to learn how to redirect output. Use ls to list the files in
your directory, saving them all to the newly created file:
% ls -l testme
-rw-rw-r-- 1 taylor
% ls -l > testme
% ls -l testme
-rw-rw-r-- 1 taylor
0 Nov 15 09:11 testme
120 Nov 15 09:12 testme
Notice that when you redirected the output, nothing was displayed on the screen;
there was no visual confirmation that it worked. But it did, as you can see by the
increased size of the new file.
3. Instead of using cat or more to view this file, try using file redirection:
% cat < testme
total 127
drwx------ 2 taylor
drwx------ 3 taylor
drwx------ 2 taylor
drwx------ 2 taylor
8
512
512
1024
512
Nov 6 14:20 Archives/
Nov 16 21:55 InfoWorld/
Nov 19 14:14 Mail/
Oct 6 09:36 News/
Filters and Piping
drwx-----drwx------rw-rw----rw-rw----rw-rw---drwx-----drwxrwx---rw-rw----
3
2
1
1
1
2
2
1
taylor
taylor
taylor
taylor
taylor
taylor
taylor
taylor
512
512
57683
46195
12556
512
512
0
Nov
Oct
Nov
Nov
Nov
Oct
Nov
Nov
11
13
20
20
16
13
8
20
10:48
10:45
20:10
06:19
09:49
10:45
22:20
20:21
147
OWL/
bin/
bitnet.lists.Z
drop.text.hqx
keylime.pie
src/
temp/
testme
8
The results are the same as if you had used the ls command, but the output file is
saved, too. You now can easily print the file or go back to it later to compare the
way it looks with the way your files look in the future.
4. Use the ls command to add some further information at the bottom of the testme
file by using >>, the append double-arrow notation:
% ls -FC >> testme
Recall that the -C flag to ls forces the system to list output in multicolumn mode.
Try redirecting the output of ls -F to a file to see what happens without the -C flag.
5. It’s time for a real-life example. You’ve finished learning UNIX, and your colleagues now consider you an expert. One afternoon, Shala tells you she has a file in
her directory, but she isn’t sure what it is. She wants to know what it is, but she
can’t figure out how to get to it. You try the file command, and UNIX tells you
the file is data. You are a bit puzzled. But then you remember file redirection:
% cat -v < mystery.file > visible.mystery.file
This command has cat -v take its input from the file mystery.file and save its
output in visible.mystery.file. All the nonprinting characters are transformed,
and Shala can poke through the file at her leisure.
Find a file on your system that file reports as a data file, and try using the redirection
commands to create a version with all characters printable through the use of cat -v.
There is an infinite number of ways that you can combine the various forms of file
redirection to create custom commands and to process files in various ways. This hour
has really just scratched the surface. Next, you learn about some popular UNIX filters and
how they can be combined with file redirection to create new versions of existing files. Also,
study the example about Shala’s file, which shows the basic steps in all UNIX file-redirection
operations: Specify the input to the command, specify the command, and specify where the
output should go.
Task 8.2: Counting Words and Lines Using wc
Writers generally talk about the length of their work in terms of number of words,
rather than number of pages. In fact, most magazines and newspapers are laid out
according to formulas based on multiplying an average-length word by the number of words
in an article.
5
Hour 8
148
These people are obsessed with counting the words in their articles, but how do they do it?
You can bet they don’t count each word themselves. If they’re using UNIX, they simply use
the UNIX wc program, which computes a word count for the file. It also can indicate the
number of characters (which ls -l indicates, too) and the number of lines in the file.
1. Start by counting the lines, words, and characters in the testme file you created
earlier in this hour:
% wc testme
4
12
% wc < testme
4
12
% cat testme | wc
4
12
121
121
121
All three of these commands offer the same result (which probably seems a bit
cryptic now). Why do you need to have three ways of doing the same thing? Later,
you learn why this is so helpful. For now, stick to using the first form of the
command.
The output is three numbers, which reveal how many lines, words, and characters,
respectively, are in the file. You can see that there are 4 lines, 12 words, and 121
characters in testme.
2. You can have wc list any one of these counts, or a combination of two, by using
different command flags: -w counts words, -c counts characters, and -l counts
lines:
% wc -w testme
12 testme
% wc -l testme
4 testme
% wc -wl testme
12
4 testme
% wc -lw testme
4
12 testme
3. Now the fun begins. Here’s an easy way to find out how many files you have in
your home directory:
% ls | wc -l
37
The ls command lists each file, one per line (because you didn’t use the -C flag).
The output of that command is fed to wc, which counts the number of lines it’s
fed. The result is that you can find out how many files you have (37) in your home
directory.
8
Filters and Piping
149
4. How about a quick gauge of how many users are on the system?
% who | wc -l
12
8
5. How many accounts are on your computer?
% cat /etc/passwd | wc -l
3877
The wc command is a great example of how the simplest of commands, when
combined in a sophisticated pipeline, can be very powerful.
Task 8.3: Removing Extraneous Lines Using uniq
Sometimes when you’re looking at a file, you’ll notice that there are many duplicate
entries, either blank lines or, perhaps, lines of repeated information. To clean up
these files and shrink their size at the same time, you can use the uniq command, which lists
each unique line in the file.
Well, it sort of lists each unique line in the file. What uniq really does is compare each line
it reads with the previous line. If the lines are the same, uniq does not list the second line. You
can use flags with uniq to get more specific results: -u lists only lines that are not repeated,
-d lists only lines that are repeated (the exact opposite of -u), and -c adds a count of how many
times each line occurred.
1. If you use uniq on a file that doesn’t have any common lines, uniq has no effect.
% uniq testme
Archives/
InfoWorld/
Mail/
News/
OWL/
bin/
bitnet.mailing-lists.Z
drop.text.hqx
keylime.pie
src/
temp/
testme
2. A trick using the cat command is that cat lists the contents of each file sequentially, even if you specify the same file over and over again, so you can easily build a
file with lots of lines:
% cat testme testme testme > newtest
Examine newtest to verify that it contains three copies of testme, one after the
other. (Try using wc.)
3. Now you have a file with duplicate lines. Will uniq realize these files have duplicate
lines? Use wc to find out:
% wc newtest
12
36
363
% uniq newtest | wc
12
36
363
5
Hour 8
150
They’re the same. Remember, the uniq command removes duplicate lines only if
they’re adjacent.
4. Create a file that has duplicate lines:
% tail -1 testme > lastline
% cat lastline lastline lastline lastline > newtest2
% cat newtest2
News/
drop.text.hqx
testme
News/
drop.text.hqx
testme
News/
drop.text.hqx
testme
News/
drop.text.hqx
testme
Now you can see what uniq does:
% uniq newtest2
News/
drop.text.hqx
testme
5. Obtain a count of the number of occurrences of each line in the file. The -c flag
does that job:
% uniq -c newtest2
4 News/
drop.text.hqx
testme
This shows that this line occurs four times in the file. Lines that are unique have no
number preface.
6. You also can see what the -d and -u flags do, and how they have exactly opposite
actions:
% uniq -d newtest2
News/
% uniq -u newtest2
%
drop.text.hqx
testme
Why did the -u flag list no output? The answer is that the -u flag tells uniq to list
only those lines that are not repeated in the file. Because the only line in the file is
repeated four times, there’s nothing to display.
Given this example, you probably think uniq is of marginal value, but you will
find that it’s not uncommon for files to have many blank lines scattered willynilly throughout the text. The uniq command is a fast, easy, and powerful way to clean up
such files.
Task 8.4: Sorting Information in a File Using sort
8
Whereas wc is useful at the end of a pipeline of commands, uniq is a filter, a program
that is really designed to be tucked in the middle of a pipeline. Filters, of course, can
be placed anywhere in a line, anywhere that enables them to help direct UNIX to do what
you want it to do. The common characteristic of all UNIX filters is that they can read input
from standard input, process it in some manner, and list the results in standard output. With
file redirection, standard input and output also can be files. To do this, you can either specify
the filenames to the command (usually input only) or use the file-redirection symbols you
learned earlier in this hour (<, >, and >>).
Filters and Piping
JUST A MINUTE
Standard input and standard output are two very common expressions in
UNIX. When a program is run, the default location for receiving input is
called standard input. The default location for output is standard output. If
you are running UNIX from a terminal, standard input and output are your
terminal.
151
8
There is a third I/O location, standard error. By default, this is the same
as standard output, but you can re-direct standard error to a different
location than standard output. You learn more about I/O redirection later
in the book.
One of the most useful filters is sort, a program that reads information and sorts it
alphabetically. You can customize the behavior of this program, like all UNIX programs, to
ignore the case of words (for example, to sort Big between apple and cat, rather than
before—most sorts put all uppercase letters before the lowercase letters), and to reverse the
order of a sort (z to a). The program sort also enables you to sort lists of numbers.
Few flags are available for sort, but they are powerful, as shown in Table 8.1.
Table 8.1. Flags for the sort command.
Flag
Function
-b
Ignore leading blanks.
Sort in dictionary order (only letters, digits, and blanks are significant).
Fold uppercase into lowercase; that is, ignore the case of words.
Sort in numerical order.
Reverse order of the sort.
-d
-f
-n
-r
1. By default, the ls command sorts the files in a directory in a case-sensitive manner.
It first lists those files that begin with uppercase letters and then those that begin
with lowercase letters:
% ls -1F
Archives/
InfoWorld/
Mail/
News/
OWL/
bin/
drop.text.hqx
5
Hour 8
152
keylime.pie
src/
temp/
testme
To force ls to list output one file per line, you can use the -1 flag (that’s
the number one, not a lowercase L).
JUST A MINUTE
To sort filenames alphabetically regardless of case, you can use sort
-f :
% ls -1 | sort -f
Archives/
bin/
drop.text.hqx
InfoWorld/
keylime.pie
Mail/
News/
OWL/
src/
temp/
testme
2. How about sorting the lines of a file? You can use the testme file you created
earlier:
% sort < testme
Archives/
InfoWorld/
Mail/
News/
OWL/
bin/
drop.text.hqx
keylime.pie
src/
temp/
testme
3. Here’s a real-life UNIX example. Of the files in your home directory, which are the
largest? The ls -s command indicates the size of each file, in blocks, and sort -n
sorts numerically:
% ls -s | sort -n
total 127
1 Archives/
1 InfoWorld/
1 Mail/
1 News/
1 OWL/
1 bin/
1 src/
1 temp/
1 testme
13 keylime.pie
46 drop.text.hqx
64 bitnet.mailing-lists.Z
8
Filters and Piping
It would be more convenient if the largest files were listed first in the output.
That’s where the -r flag to reverse the sort order can be useful:
% ls -s | sort -nr
64 bitnet.mailing-lists.Z
46 drop.text.hqx
13 keylime.pie
1 testme
1 temp/
1 src/
1 bin/
1 OWL/
1 News/
1 Mail/
1 InfoWorld/
1 Archives/
total 127
153
8
4. One more refinement is available to you. Instead of listing all the files, use the head
command, and specify that you want to see only the top five entries:
% ls
64
46
13
1
1
-s | sort -nr | head -5
drop.text.hqx
keylime.pie
testme
temp/
That’s a powerful and complex UNIX command, yet it is composed of simple and
easy-to-understand components.
Like many of the filters, sort isn’t too exciting by itself. As you explore UNIX further
and learn more about how to combine these simple commands to build sophisticated
instructions, you will begin to see their true value.
Task 8.5: Number Lines in Files Using cat -n and nl
It often can be helpful to have a line number listed next to each line of a file. It’s quite
simple to do with the cat program by specifying the -n flag to number lines in the
file displayed.
On many UNIX systems, there’s a considerably better command for numbering lines in a file
and for many other tasks. The command nl, for number lines, is an AT&T System V command. A system that doesn’t have the nl command will complain nl: command not found.
If you have this result, experiment with cat -n instead.
Step 2. Action
1. Because one of my own systems did not have the nl command, I moved to one that
had the nl command for this example. I quickly rebuilt the testme file:
% ls -l > testme
5
Hour 8
154
To see line numbers now, cat
% cat -n testme
1 total 60
2 -rw-r--r-3 -rw------4 drwx-----5 drwxr-xr-x
6 drwxr-xr-x
7 drwxr-xr-x
8 -rw-r--r-9 -rw-r--r--
1
1
4
2
2
2
1
1
-n
taylor
taylor
taylor
taylor
taylor
taylor
taylor
taylor
will work fine:
1861
22194
4096
4096
4096
4096
12445
0
Jun
Oct
Nov
Nov
Nov
Nov
Sep
Nov
2
1
13
13
13
13
17
20
1992
1992
11:09
11:09
11:09
11:09
14:56
18:16
Global.Software
Interactive.Unix
Mail/
News/
Src/
bin/
history.usenet.Z
testme
2. The alternative, which does exactly the same thing here, is to try nl without any
flags:
% nl testme
1 total 60
2 -rw-r--r-3 -rw------4 drwx-----5 drwxr-xr-x
6 drwxr-xr-x
7 drwxr-xr-x
8 -rw-r--r-9 -rw-r--r--
1
1
4
2
2
2
1
1
taylor
taylor
taylor
taylor
taylor
taylor
taylor
taylor
1861
22194
4096
4096
4096
4096
12445
0
Jun
Oct
Nov
Nov
Nov
Nov
Sep
Nov
2
1
13
13
13
13
17
20
1992
1992
11:09
11:09
11:09
11:09
14:56
18:16
Global.Software
Interactive.Unix
Mail/
News/
Src/
bin/
history.usenet.Z
testme
3. Notice that both commands also can number lines fed to them via a command
pipeline:
% ls -CF | cat -n
1 Global.Software
2 Interactive.Unix
3 Mail/
1 Global.Software
2 Interactive.Unix
3 Mail/
News/
Src/
bin/
News/
Src/
bin/
history.usenet.Z
testme
% ls -CF | nl
history.usenet.Z
testme
Like many other UNIX tools, nl and its doppelganger cat -n aren’t very thrilling by
themselves. As additional members in the set of powerful UNIX tools, however, they
can prove tremendously helpful in certain situations. As you soon will see, nl also has some
powerful options that can make it a bit more fun.
Task 8.6: Cool nl Tricks and Capabilities
A program that prefaces each line with a line number isn’t much of an addition to
the UNIX command toolbox, so the person who wrote the nl program added some
further capabilities. With different command flags, nl can either number all lines (by default
it numbers only lines that are not blank) or skip line numbering (which means it’s an
additional way to display the contents of a file). The best option, though, is that nl can
selectively number just those lines that contain a specified pattern.
8
Filters and Piping
JUST A MINUTE
155
If you don’t have the nl command on your system, I’m afraid you’re out of
luck in this section. Later in the book, you learn other ways to accomplish
these tasks. For now, though, if you don’t have nl, skip to the next hour
and start to learn about the grep command.
8
The command flag format for nl is a bit more esoteric than you’ve seen up to this point. The
different approaches to numbering lines with nl are all modifications of the -b flag (for body
numbering options). The four flags are -ba, which numbers all lines; -bt, which numbers
printable text only; -bn, which results in no numbering; and -bp pattern, for numbering lines
that contain the specified pattern.
One final option is to insert a different separator between the line number and the line by
telling nl to use -s, the separator flag.
1. To begin, I’ll use a command that you haven’t seen before to add a few blank lines
to the testme file. The echo command simply writes back to the screen anything
specified. Try echo hello.
% rm testme
% ls -CF > testme
% echo “” >> testme
% echo “” >> testme
% ls -CF >> testme
% cat testme
Global.Software
Interactive.Unix
Mail/
News/
Src/
bin/
history.usenet.Z
testme
Global.Software
Interactive.Unix
Mail/
News/
Src/
bin/
history.usenet.Z
testme
JUST A MINUTE
Parts of UNIX are rather poorly designed, as you have already learned.
For example, if you use the echo command without arguments, you get no
output. However, if you add an empty argument (a set of quotation marks
with nothing between them), echo outputs a blank line. It doesn’t make
much sense, but it works.
5
Hour 8
156
2. Now watch what happens when nl uses its default settings to number the lines in
testme:
% nl testme
1 Global.Software
2 Interactive.Unix
3 Mail/
4
5
6
Global.Software
Interactive.Unix
Mail/
News/
Src/
bin/
history.usenet.Z
testme
News/
Src/
bin/
history.usenet.Z
testme
You can accomplish the same thing by specifying nl -bt testme. Try this to verify
that your system gives the same results.
3. It’s time to use one of the new two-letter command options to number the lines,
including the blank lines:
% nl -ba testme
1 Global.Software
2 Interactive.Unix
3 Mail/
4
5
6 Global.Software
7 Interactive.Unix
8 Mail/
News/
Src/
bin/
history.usenet.Z
testme
News/
Src/
bin/
history.usenet.Z
testme
4. If you glance at the contents of my testme file, you can see that two lines contain
the word history. To have nl number just those lines, try the -bp pattern-matching
option:
% nl -bphistory testme
1 Global.Software
News/
Interactive.Unix Src/
Mail/
bin/
2
Global.Software
News/
Mail/
bin/
history.usenet.Z
testme
history.usenet.Z
testme
Notice that numbering the two lines has caused the rest of the lines to fall out of
alignment on the display.
5. This is when the -s, or separator, option comes in handy:
% nl -bphistory -s: testme
1:Global.Software
News/
Mail/
bin/
2:Global.Software
News/
Mail/
bin/
8
history.usenet.Z
testme
history.usenet.Z
testme
Filters and Piping
157
In this case, I specified that instead of using a tab, which is the default separator
between the number and line, nl should use a colon. As you can see, the output
now lines up again.
Just about anything can be specified as the separator, as sensible or weird as it
might be:
% nl -s’, line is: ‘ testme
1, line is: Global.Software
history.
usenet.Z
2, line is: Interactive.Unix
3, line is: Mail/
4, line is: Global.Software
history.
usenet.Z
5, line is: Interactive.Unix
6, line is: Mail/
8
News/
Src/
bin/
testme
News/
Src/
bin/
testme
Notice the use of single quotation marks (‘) in this example. I want to include
spaces as part of my pattern, so I need to ensure that the program knows this. If I
didn’t use the quotation marks, nl would use a comma as the separator and then
tell me that it couldn’t open a file called line or is:.
The nl command demonstrates that there are plenty of variations on simple
commands. When you read earlier that you would learn how to number lines in a file,
did you think that this many subtleties were involved?
Summary
You have learned quite a bit in this hour and are continuing down the road to UNIX expertise.
You learned about file redirection. You can’t go wrong by spending time studying these
closely. The concept of using filters and building complex commands by combining simple
commands with pipes has been more fully demonstrated here, too. This higher level of UNIX
command language is what makes UNIX so powerful and easy to mold.
This hour hasn’t skimped on commands, either. It introduced wc for counting lines, words,
and characters in a file (or more than one file: try wc * in your home directory). You also
learned to use the uniq, sort, and spell commands. You learned about using nl for
numbering lines in a file—in a variety of ways—and cat -n as an alternative “poor person’s”
line-numbering strategy. You also were introduced to the echo command.
By the way, the echo command also can tell you about specific environment variables, just
like env or printenv do. Try echo $HOME or echo $PATH to see what happens, and compare the
output with env HOME and env PATH.
5
Hour 8
158
Workshop
in the next hour.
Key Terms
file redirection Most UNIX programs expect to read their input from the user (that is,
standard input) and write their output to the screen (standard output). By use of file
redirection, however, input can come from a previously created file, and output can be saved
to a file instead of being displayed on the screen.
filter Filters are a particular type of UNIX program that expects to work either with file
redirection or as part of a pipeline. These programs read input from standard input, write
output to standard output, and often don’t have any starting arguments.
standard input UNIX programs always default to reading information from the user by
reading the keyboard and watching what’s typed. With file redirection, input can come from
a file, and with pipelines, input can be the result of a previous UNIX command.
standard error This is the same as standard output, but you can redirect standard error to
a different location than standard output.
standard output When processing information, UNIX programs default to displaying the
output on the screen itself, also known as standard output. With file redirection, output can
easily be saved to a file; with pipelines, output can be sent to other programs.
Questions
1. The placement of file-redirection characters is important to ensure that the command works correctly. Which of the following six commands do you think will
work, and why?
< file wc
cat file | wc
wc file <
cat < file | wc
wc < file
wc | cat
Now try them and see if you’re correct.
2. The wc command can be used for lots of different tasks. Try to imagine a few that
would be interesting and helpful to learn (for example, how many users are on the
system right now?). Try them on your system.
3. Does the file size listed by wc -c always agree with the file size listed by the ls
command? With the size indicated by ls -s? If there is any difference, why?
8
Filters and Piping
4. What do you think would happen if you tried to sort a list of words by pretending
they’re all numbers? Try it with the command ls -1 | sort -n to see what
happens. Experiment with the variations.
5. Do you spell your filenames correctly? Use spell to find out.
159
8
The next hour introduces wildcards and regular expressions, and tools to use those powerful
concepts. You learn how these commands can help you extract data from even the most
unwieldy files.
You learn one of the secret UNIX commands for those really in the know, the secret-society,
pattern-matching program grep. Better yet, you learn how it got its weird and confusing
name! You also learn about the tee command and the curious-but-helpful << file-redirection
command.
5
Hour
9
Wildcards and Regular
Expressions
One of the trickiest aspects of UNIX is the concept of wildcards and regular
expressions. Wildcards are a tool that allows you to “guess” at a filename, or to
specify a group of filenames easily. Regular expressions are pattern-matching
tools that are different, and more powerful, than wildcards.
You’ll meet two new commands, sed and grep, that use regular expressions.
Goals for This Hour
In this hour, you learn about
■
■
■
■
■
■
■
Filename wildcards
Advanced wildcards
Regular expressions
Searching files using grep
A more powerful grep
A fast grep
Using the stream editor sed to change output on-the-fly
161
9
Hour 9
162
This hour begins by looking at the two pattern-matching tools frequently found in UNIX.
A foray into commands that use these tools immediately follows.
Task 9.1: Filename Wildcards
By now you are doubtless tired of typing every letter of each filename into your
system for each example. There is a better and easier way! Just as the special card in
poker can have any value, UNIX has special characters that the various shells (the commandline interpreter programs) all interpret as wildcards. This allows for much easier typing of
patterns.
There are two wildcards to learn here: * acts as a match for any number and sequence of
characters, and ? acts as a match for any single character. In the broadest sense, a lone * acts
as a match for all files in the current directory (in other words, ls * is identical to ls), whereas
a single ? acts as a match for all one-character-long filenames in a directory (for instance, ls
?, which will list only those filenames that are one character long). The following examples
will make this clear.
1. Start by using ls to list your home directory.
% ls -CF
Archives/
InfoWorld/
Mail/
News/
OWL/
bin/
drop.text.hqx
keylime.pie
src/
temp/
testme
2. To experiment with wildcards, it’s easiest to use the echo command. If you recall,
echo repeats anything given to it, but—and here’s the secret to its value—the shell
interprets anything that is entered before the shell lets echo see it. That is, the * is
expanded before the shell hands the arguments over the command.
% echo *
Archives InfoWorld Mail News OWL bin bitnet.mailing-lists.Z
drop.text.hqx keylime.pie src temp testme
Using the * wildcard enables me to reference easily all files in the directory. This is
quite helpful.
3. A wildcard is even more helpful than the example suggests because it can be
embedded in the middle of a word or otherwise used to limit the number of
matches. To see all files that began with the letter t, use the *:
% echo t*
temp testme
Try echo b* to see all your files that start with the letter b.
9
163
4. Variations are possible, too. I could use wildcards to list all files or directories that
end with the letter s:
% echo *s
Archives News
Watch what happens if I try the same command using the ls command rather than
the echo command:
% ls -CF *s
Archives:
Interleaf.story
Opus.story
Tartan.story.Z
interactive.txt.Z
News:
mailing.lists.usenet
usenet.1
nextstep.txt.Z
rae.assist.infoworld.Z
usenet.alt
Using the ls command here makes UNIX think I want it to list two directories,
not just the names of the two files. This is where the -d flag to ls could prove
helpful to force a listing of the directories rather than of their contents.
5. Notice that, in the News directory, I have three files with the word usenet somewhere in their names. The wildcard pattern usenet* would match two of them, and
*usenet would match one. A valuable aspect of the * wildcard is that it can match
zero or more characters, so the pattern *usenet* will match all three.
% echo News/*usenet*
News/mailing.lists.usenet News/usenet.1 News/usenet.alt
Also notice that wildcards can be embedded in a filename or pathname. In this
example, I specified that I was interested in files in the News directory.
6. Could you match a single character? To see how this can be helpful, it’s time to
move into a different directory, OWL on my system.
% cd OWL
% ls -CF
Student.config
WordMap/
owl*
owl.c
owl.data
owl.h
owl.o
simple.editor.c
simple.editor.o
If I request owl*, which files will be listed?
% echo owl*
owl owl.c owl.data owl.h owl.o
What do I do if I am interested only in the source, header, and object files, which
are here indicated by a .c, .h, or .o suffix. Using a wildcard that matches zero or
more letters won’t work; I don’t want to see owl or owl.data. One possibility would
be to use the pattern owl.* (by adding the period, I can eliminate the owl file itself).
What I really want, however, is to be able to specify all files that start with the four
characters owl. and have exactly one more character. This is a situation in which
the ? wildcard works:
9
Hour 9
164
% echo owl.?
owl.c owl.h owl.o
Because no files have exactly one letter following the three letters owl, watch what
happens when I specify owl? as the pattern:
% echo owl?
echo: No match.
This leads to a general observation. If you want to have echo return a question to
you, you have to do it carefully because the shell interprets the question mark as a
wildcard:
% echo are you listening?
echo: No match.
To accomplish this, you simply need to surround the entire question with quotation marks:
% echo ‘are you listening?’
are you listening?
It won’t surprise you that there are more complex ways of using wildcards to build
filename patterns. What likely will surprise you is that the vast majority of UNIX
users don’t even know about the * and ? wildcards! This knowledge gives you a definite
advantage.
Task 9.2: Advanced Filename Wildcards
Earlier, you learned about two special wildcard characters that can help you when
specifying files for commands in UNIX. The first was the ?, which matches any single
character, and the other was the *, which matches zero or more characters. There are more
special wildcards for the shell when specifying filenames, and it’s time to learn about another
of them.
This new notation is known as a character range, serving as a wildcard less general than the
question mark.
1. A pair of square brackets denotes a range of characters, which can be either explicitly listed or indicated as a range with a dash between them. I’ll start with a list
of files in my current directory:
% ls
Archives/
InfoWorld/
Mail/
9
News/
OWL/
awkscript
bigfiles
bin/
keylime.pie
owl.c
sample
sample2
src/
temp/
tetme
165
If I want to see both bigfiles and the bin directory, I can use b* as a file pattern:
% ls -ld b*
-rw-rw---- 1 taylor
drwx------ 2 taylor
165 Dec 3 16:42 bigfiles
512 Oct 13 10:45 bin/
If I want to see all entries that start with a lowercase letter, I can explicitly type each
one:
% ls -ld a*
-rw-rw----rw-rw---drwx------rw-rw----rw-rw----rw-rw----rw-rw---drwx-----drwxrwx---rw-rw----
b* k* o* s* t*
1 taylor
126 Dec 3 16:34 awkscript
1 taylor
165 Dec 3 16:42 bigfiles
2 taylor
512 Oct 13 10:45 bin/
1 taylor
12556 Nov 16 09:49 keylime.pie
1 taylor
8729 Dec 2 21:19 owl.c
1 taylor
199 Dec 3 16:11 sample
1 taylor
207 Dec 3 16:11 sample2
2 taylor
512 Oct 13 10:45 src/
2 taylor
512 Nov 8 22:20 temp/
1 taylor
582 Nov 27 18:29 tetme
That’s clearly quite awkward. Instead, I can specify a range of characters to match.
I specify the range by listing them all tucked neatly into a pair of square brackets:
% ls -ld [abkost]*
-rw-rw---- 1 taylor
-rw-rw---- 1 taylor
drwx------ 2 taylor
-rw-rw---- 1 taylor
-rw-rw---- 1 taylor
-rw-rw---- 1 taylor
-rw-rw---- 1 taylor
drwx------ 2 taylor
drwxrwx--- 2 taylor
-rw-rw---- 1 taylor
126
165
512
12556
8729
199
207
512
512
582
Dec
Dec
Oct
Nov
Dec
Dec
Dec
Oct
Nov
Nov
3
3
13
16
2
3
3
13
8
27
16:34
16:42
10:45
09:49
21:19
16:11
16:11
10:45
22:20
18:29
awkscript
bigfiles
bin/
keylime.pie
owl.c
sample
sample2
src/
temp/
tetme
In this case, the shell matches all files that start with an a, b, k, o, s, or t. This
notation is still a bit clunky and would be more so if there were more files involved.
2. The ideal is to specify a range of characters by using the hyphen character in the
middle of a range:
% ls -ld [a-z]*
-rw-rw---- 1 taylor
-rw-rw---- 1 taylor
drwx------ 2 taylor
-rw-rw---- 1 taylor
-rw-rw---- 1 taylor
-rw-rw---- 1 taylor
-rw-rw---- 1 taylor
drwx------ 2 taylor
drwxrwx--- 2 taylor
-rw-rw---- 1 taylor
126
165
512
12556
8729
199
207
512
512
582
Dec
Dec
Oct
Nov
Dec
Dec
Dec
Oct
Nov
Nov
3
3
13
16
2
3
3
13
8
27
16:34
16:42
10:45
09:49
21:19
16:11
16:11
10:45
22:20
18:29
awkscript
bigfiles
bin/
keylime.pie
owl.c
sample
sample2
src/
temp/
tetme
In this example, the shell will match any file that begins with a lowercase letter,
ranging from a to z, as specified.
9
Hour 9
166
3. Space is critical in all wildcard patterns, too. Watch what happens if I accidentally
add a space between the closing bracket of the range specification and the asterisk
following:
% ls -CFd [a-z] *
Archives/
News/
InfoWorld/
OWL/
Mail/
awkscript
bigfiles
bin/
keylime.pie
owl.c
sample
sample2
src/
temp/
tetme
This time, the shell tried to match all files whose names were one character long
and lowercase, and then it tried to match all files that matched the asterisk
wildcard, which, of course, is all regular files in the directory.
4. The combination of character ranges, single-character wildcards, and multicharacter wildcards can be tremendously helpful. If I move to another directory, I
can easily search for all files that contain a single digit, dot, or underscore in the
name:
% cd Mail
% ls -CF
71075.446
72303.2166
bmcinern
bob_gull
cennamo
dan_some
dataylor
decc
disserli
druby
dunlaplm
ean_huts
emilyc
gordon_hat
harrism
j=taylor
james
jeffv
john_welch
john_prage
kcs
lehman
lenz
mac
mailbox
manley
mark
marmi
marv
matt_ruby
mcwillia
netnews.postings
raf
rexb
rock
rustle
sartin
sent
shalini
siob_n
steve
tai
taylor
v892127
wcenter
windows
xd1f
% ls *[0-9._]*
71075.446
72303.2166
bob_gull
dan_some
ean_huts
gordon_hat
john_welcher
john_prage
matt_ruby
netnews.postings
siob_n
v892127
xd1f
I think that the best way to learn about pervasive features of UNIX such as shell
filename wildcards is just to use them. If you flip through this book, you immediately
notice that the examples are building on earlier information. This will continue to be the case,
and the filename range notation shown here will be used again and again, in combination
with the asterisk and question mark, to specify groups of files or directories.
Remember that, if you want to experiment with filename wildcards, you can most easily use
the echo command because it dutifully prints the expanded version of any pattern you specify.
9
167
Task 9.3: Creating Sophisticated Regular
Expressions
A regular expression can be as simple as a word to be matched letter for letter, such
as acme, or as complex as the example in the printers script, ‘(^[a-zA-Z]|:wi)’,
which matches all lines that begin with an upper- or lowercase letter or that contain :wi.
The language of regular expressions is full of punctuation characters and other letters used in
unusual ways. It is important to remember that regular expressions are different from shell
wildcard patterns. It’s unfortunate, but it’s true. In the C shell, for example, a* lists any file
that starts with the letter a. Regular expressions aren’t left rooted, which means that you need
to specify â if you want to match only lines that begin with the letter a. The shell pattern
a* matches only filenames that start with the letter a, and the * has a different interpretation
completely when used as part of a regular expression: a* is a pattern that matches zero or more
occurrences of the letter a. The notation for regular expressions is shown in Table 9.1. The
egrep command has additional notation that you will learn shortly.
Table 9.1. Summary of regular-expression notation.
Notation
Meaning
c
Matches the character c
Forces c to be read as the letter c, not as another meaning the character
might have
Beginning of the line
End of the line
Any single character
Any single character in the set specified
Any single character not in the set specified
Zero or more occurrences of character c
\c
^
$
.
[xy]
[^xy]
c*
The notation isn’t as complex as it looks in this table. The most important things to remember
about regular expressions are that the * denotes zero or more occurrences of the previous
character, and . is any single character. Remember that shell patterns use * to match any set
of zero or more characters independent of the previous character, and ? to match a single
character.
9
168
Hour 9
1. The easy searches with grep are those that search for specific words without any
special regular expression notation:
taylorj:?:1048:1375:James Taylor:/users/taylorj:/bin/csh
mtaylor:?:769:1375:Mary Taylor:/users/mtaylor:/usr/local/bin/tcsh
dataylor:?:375:518:Dave Taylor:/users/dataylor:/usr/local/lib/msh
taylorjr:?:203:1022:James Taylor:/users/taylorjr:/bin/csh
taylorrj:?:662:1042:Robert Taylor:/users/taylorrj:/bin/csh
taylorm:?:869:1508:Melanie Taylor:/users/taylorm:/bin/csh
taylor:?:1989:1412:Dave Taylor:/users/taylor:/bin/csh
I searched for all entries in the passwd file that contain the pattern taylor.
2. I’ve found more matches than I wanted, though. If I’m looking for my own
account, I don’t want to see all these alternatives. Using the ^ character before the
pattern left-roots the pattern:
% grep “^taylor” /etc/passwd
taylorm:?:869:1508:Melanie Taylor:/users/taylorm:/bin/csh
taylor:?:1989:1412:Dave Taylor:/users/taylor:/bin/cshx
Now I want to narrow the search further. I want to specify a pattern that says
“show me all lines that start with taylor, followed by a character that is not a
lowercase letter.”
3. To accomplish this, I use the [^xy] notation, which indicates an exclusion set, or set
of characters that cannot match the pattern:
% grep “^taylor[â-z]” /etc/passwd
It worked! You can specify a set two ways: You can either list each character or use
a hyphen to specify a range starting with the character to the left of the hyphen and
ending with the character to the right of the hyphen. That is, a-z is the range
beginning with a and ending with z, and 0-9 includes all digits.
4. To see which accounts were excluded, remove the ^ to search for an inclusion range,
which is a set of characters of which one must match the pattern:
% grep ‘^taylor[a-z]’ /etc/passwd
taylormx:?:869:1508:Melanie Taylor:/users/taylorm:/bin/csh
9
169
5. To see some other examples, I use head to view the first 10 lines of the password file:
% head /etc/passwd
root:?:0:0:root:/:/bin/csh
actmaint:?:23:809:Maintenance:/usr/adm/actmaint:/bin/ksh
pop:*?:26:819::/usr/spool/pop:/bin/csh
lp:*?:70:10:Lp Admin:/usr/spool/lp:
trouble:*?:97:501:Report Facility:/usr/mrg/trouble:/usr/local/lib/msh
Now I’ll specify a pattern that tells grep to search for all lines that contain zero or
more occurrences of the letter z.
% grep ‘z*’ /etc/passwd | head
root:?:0:0:root:/:/bin/csh
actmaint:?:23:809:Maintenance:/usr/adm/actmaint:/bin/ksh
pop:*?:26:819::/usr/spool/pop:/bin/csh
lp:*?:70:10:Lp Adminuniverse(att):/usr/spool/lp:
trouble:*?:97:501:Report Facility:/usr/mrg/trouble:/usr/local/lib/msh
Broken pipe
The result is identical to the previous command, but it shouldn’t be a surprise.
Specifying a pattern that matches zero or more occurrences will match every line!
Specifying only the lines that have one or more z’s produces output that is a bit
more odd looking:
% grep ‘zz*’ /etc/passwd | head
marg:?:724:1233:Guyzee:/users/marg:/bin/ksh
axy:?:1272:1233:martinez:/users/axy:/bin/csh
wizard:?:1560:1375:Oz:/users/wizard:/bin/ksh
zhq:?:2377:1318:Zihong:/users/zhq:/bin/csh
mm:?:7152:1233:Michael Kenzie:/users/mm:/bin/ksh
tanzm:?:7368:1140:Zhen Tan:/users/tanzm:/bin/csh
mendozad:?:8176:1233:Don Mendoza:/users/mendozad:/bin/csh
pavz:?:8481:1175:Mary L. Pavzky:/users/pavz:/bin/csh
hurlz:?:9189:1375:Tom Hurley:/users/hurlz:/bin/csh
tulip:?:9222:1375:Liz Richards:/users/tulip:/bin/csh
Broken pipe
6. Earlier I found that a couple lines in the /etc/passwd file were for accounts that
didn’t specify a login shell. Each line in the password file must have a certain
number of colons, and the very last character on the line for these accounts will be
a colon, an easy grep pattern:
% grep ‘:$’ /etc/passwd
lp:*?:70:10:System V Lp Adminuniverse(att):/usr/spool/lp:
9
Hour 9
170
7. Consider this. I get a call from my accountant, and I need to find a file containing
a message about a $100 outlay of cash to buy some software. I can use grep to
search for all files that contain a dollar sign, followed by a one, followed by one or
more zeroes:
% grep ‘$100*’ * */*
Mail/bob_gale:
Unfortunately, our fees are currently $100 per test
➥drive, budgets
Mail/dan_sommer:We also pay $100 for Test Drives, our very short “First
➥Looks” section. We often
Mail/james:has been dropped, so if I ask for $1000 is that way outta
➥line
Mail/john_spragens:time testing things since it’s a $100 test drive: I’m
➥willing to
Mail/john_spragens:
Finally, I’d like to request $200 rather than
➥$100 for
Mail/mac:again: expected pricing will be $10,000 - $16,000 and the
➥BriteLite LX with
Mail/mark:I’m promised $1000 / month for a first
Mail/netnews.postings: Win Lose or Die, John Gardner (hardback) $10
Mail/netnews.postings:I’d be willing to pay, I dunno, $100 / year for
➥the space? I would
Mail/sent:to panic that they’d want their $10K advance back, but the
➥good news is
Mail/sent:That would be fine. How about $100 USD for both, to include
➥any
Mail/sent:
Amount: $100.00
That’s quite a few matches. Notice that among the matches are $1000, $10K, and
$10. To match the specific value $100, of course, I can use $100 as the search
pattern.
TIME SAVER
You can use the shell to expand files not just in the current directory, but
one level deeper into subdirectories, too: * expands your search beyond
files in the current directory, and */* expands your search to all files
contained one directory below the current point. If you have lots of files,
you might instead see the error arg list too long; that’s where the find
command proves handy.
This pattern demonstrates the sophistication of UNIX with regular expressions.
For example, the $ character is a special character that can be used to indicate the
end of a line, but only if it is placed at the very end of the pattern. Because I did
not place it at the end of the pattern, the grep program reads it as the $ character
itself.
8. Here’s one more example. In the old days, when people were tied to typewriters, an
accepted convention for writing required that you put two spaces after the period at
the end of a sentence even though only one space followed the period of an
9
abbreviation such as J. D. Salinger. Nowadays, with more text being produced
through word processing and desktop publishing, the two-space convention is less
accepted, and indeed, when submitting work for publication, I often have to be
sure that I don’t have two spaces after punctuation lest I get yelled at! The grep
command can help ferret out these inappropriate punctuation sequences, fortunately; but the pattern needed is tricky.
To start, I want to see if, anywhere in the file dickens.note, I have used a period
followed by a single space:
% grep ‘. ‘ dickens.note
A Tale of Two Cities
Preface
When I was acting, with my children and friends, in Mr Wilkie Collins’s
drama of The Frozen Deep, I first conceived the main idea of this
story. A strong desire came upon me then, to
embody it in my own person;
and I traced out in my fancy, the state of mind of which it would
necessitate the presentation
to an observant spectator, with particular
care and interest.
As the idea became familiar to me, it gradually shaped itself into its
present form. Throughout its execution, it has had complete possession
of me; I have so far verified what
is done and suffered in these pages,
as that I have certainly done and suffered it all myself.
Whenever any reference (however slight) is made here to the condition
of the Danish people before or during the Revolution, it is truly made,
on the faith of the most trustworthy
witnesses. It has been one of my hopes to add
something to the popular and picturesque means of
understanding that terrible time, though no one can hope
to add anything to the philosophy of Mr Carlyle’s wonderful book.
Tavistock House
November 1859
What’s happening here? The first line doesn’t have a period in it, so why does grep
say it matches the pattern? In grep, the period is a special character that matches
any single character, not specifically the period itself. Therefore, my pattern
matches any line that contains a space preceded by any character.
To avoid this interpretation, I must preface the special character with a backslash
(\) if I want it to be read as the . character itself:
% grep ‘\. ‘ dickens.note
present form. Throughout its execution, it has had complete posession
Ahhh, that’s better. Notice that all three of these lines have two spaces after each period.
With the relatively small number of notations available in regular expressions, you
can create quite a variety of sophisticated patterns to find information in a file.
171
9
Hour 9
172
Task 9.4: Searching Files Using grep
Two commonly used commands are the key to you becoming a power user and
becoming comfortable with the capabilities of the system. The ls command is one
example, and the grep command is another. The oddly named grep command makes it easy
to find lost files or to find files that contain specific text.
JUST A MINUTE
After laborious research and countless hours debating with UNIX developers, I am reasonably certain that the derivation of the name grep is as
follows: Before this command existed, UNIX users would use a crude linebased editor called ed to find matching text. As you know, search patterns
in UNIX are called regular expressions. To search throughout a file, the
user prefixed the command with global. Once a match was made, the
user wanted to have it either listed to the screen with print. To put it all
together, the operation was global/regular expression/print. That
phrase was pretty long, however, so users shortened it to g/re/p. Thereafter, when a command was written, grep seemed to be a natural, if odd
and confusing, name.
The grep command not only has a ton of different command options, but it has two
variations in UNIX systems, too. These variations are egrep, for specifying more complex
patterns (regular expressions), and fgrep, for using file-based lists of words as search patterns.
You could spend the next 100 pages learning all the obscure and weird options to the grep
family of commands. When you boil it down, however, you’re probably going to use only
the simplest patterns and maybe a useful flag or two. Think of it this way: Just because there
are more than 500,000 words in the English language (according to the Oxford English
Dictionary) doesn’t mean that you must learn them all to communicate effectively.
With this in mind, youl learn the basics of grep this hour, but you’ll pick up more insight
into the program’s capabilities and options during the next few hours.
A few of the most important grep command flags are listed in Table 9.2.
Table 9.2. The most helpful grep flags.
Flag
Function
-c
List a count of matching lines only.
Ignore the case of the letters in the pattern.
List filenames of files that match the specified pattern only.
Include line numbers.
-i
-l
-n
9
173
1. Begin by making sure you have a test file to work with. The example shows the
testme file from the previous uniq examples:
% cat testme
Archives/
InfoWorld/
Mail/
News/
OWL/
bin/
drop.text.hqx
keylime.pie
src/
temp/
testme
2. The general form of grep is to specify the command, any flags you want to add, the
pattern, and a filename:
% grep bitnet testme
Mail/
temp/
As you can see, grep easily pulled out the line in the testme file that contained the
pattern bitnet.
3. Be aware that grep finds patterns in a case-sensitive manner:
% grep owl testme
%
Note that OWL was not found because the pattern specified with the grep command
was all lowercase, owl.
But that’s where the -i flag can be helpful, which causes grep to ignore case:
% grep -i owl testme
Archives/
OWL/
keylime.pie
4. For the next few examples, I’ll move into the /etc directory because some files
therein there have lots of lines. The wc command shows that the file /etc/passwd
has almost 4,000 lines:
% cd /etc
% wc -l /etc/passwd
3877
My account is taylor. I’ll use grep to see my account entry in the password file:
mtaylor:?:760:1375:Mary Taylor:/users/mtaylor:/usr/local/bin/tcsh
dataylor:?:375:518:Dave Taylor:/users/dataylor:/usr/local/lib/msh
taylorm:?:862:1508:Melanie Taylor:/users/taylormx:/bin/csh
Try this on your system, too.
9
Hour 9
174
5. As you can see, many accounts contain the pattern taylor.
A smarter way to see how often the taylor pattern appears is to use the -c flag to
grep, which will indicate how many case-sensitive matches are in the file before any
of them are displayed on the screen:
% grep -c taylor /etc/passwd
7
The command located seven matches. Count the listing in instruction 4 to confirm
this.
6. With 3,877 lines in the password file, it could be interesting to see if all the Taylors
started their accounts at about the same time. (This presumably would mean they
all appear in the file at about the same point.) To do this, I’ll use the -n flag to
number the output lines:
% grep -n taylor /etc/passwd
319:taylorj:?:1048:1375:James Taylor:/users/taylorj:/bin/csh
1314:mtaylor:?:760:1375:Mary Taylor:/users/mtaylor:/usr/local/bin/tcsh
1419:dataylor:?:375:518:Dave Taylor:/users/dataylor:/usr/local/lib/msh
1547:taylorjr:?:203:1022:James Taylor:/users/taylorjr:/bin/csh
1988:taylorrj:?:668:1042:Robert Taylor:/users/taylorrj:/bin/csh
2133:taylorm:?:8692:1508:Melanie Taylor:/users/taylorm:/bin/csh
3405:taylor:?:1989:1412:Dave Taylor:/users/taylor:/bin/csh
This is a great example of a default separator adding incredible confusion to the
output of a command. Normally, a line number followed by a colon would be no
problem, but in the passwd file (which is already littered with colons), it’s confusing. Compare this output with the output obtained in instruction 4 with the grep
command alone to see what’s changed.
You can see that my theory about when the Taylors started their accounts was
wrong. If proximity in the passwd file is an indicator that accounts are assigned at
similar times, then no Taylors started their accounts even within the same week.
These examples of how to use grep barely scratch the surface of how this powerful and
sophisticated command can be used. Explore your own file system using grep to
search files for specific patterns.
JUST A MINUTE
9
Armed with wildcards, you now can try the -l flag to grep, which, as you
recall, indicates the names of the files that contain a specified pattern,
rather than printing the lines that match the pattern. If I go into my
electronic mail archive directory—Mail—I can easily, using the command
grep -l -i chicago Mail/*, search for all files that contain Chicago. Try
using grep -l to search across all files in your home directory for words
or patterns.
175
Task 9.5: For Complex Expressions, Try egrep
Sometimes a single regular expression can’t locate what you seek. For example,
perhaps you’re looking for lines that have either one pattern or a second pattern.
That’s where the egrep command proves helpful. The command gets its name from
“expression grep,” and it has a notational scheme more powerful than that of grep, as shown
in Table 9.3.
Table 9.3. Regular expression notation for egrep .
Notation
Meaning
c
Matches the character c
Forces c to be read as the letter c, not as another meaning the character
might have
Beginning of the line
End of the line
Any single character
Any single character in the set specified
Any single character not in the set specified
Zero or more occurrences of character c
One or more occurrences of character c
Zero or one occurrences of character c
Either a or b
Regular expression
\c
^
$
.
[xy]
[^xy]
c*
c+
c?
a|b
(a)
1. Now I’ll search the password file to demonstrate egrep. A pattern that seemed a bit
weird was the one used with grep to search for lines containing one or more
occurrences of the letter z: ‘zz*’. With egrep, this search is much easier:
% egrep ‘z+’ /etc/passwd | head
axy:?:1272:1233:martinez:/users/axy:/bin/csh
zhq:?:2377:1318:Zihong:/users/zhq:/bin/csh
mm:?:7152:1233:Michael Kenzie:/users/mm:/bin/ksh
tanzm:?:7368:1140:Zhen Tan:/users/tanzm:/bin/csh
mendozad:?:8176:1233:Don Mendoza:/users/mendozad:/bin/csh
pavz:?:8481:1175:Mary L. Pavzky:/users/pavz:/bin/csh
hurlz:?:9189:1375:Tom Hurley:/users/hurlz:/bin/csh
tulip:?:9222:1375:Liz Richards:/users/tulip:/bin/csh
Broken pipe
9
176
Hour 9
2. To search for lines that have either a z or a q, I can use the following:
% egrep ‘(z|q)’ /etc/passwd | head
aaq:?:528:1233:Don Kid:/users/aaq:/bin/csh
abq:?:560:1233:K Laws:/users/abq:/bin/csh
ahq:?:752:1233:Andy Smith:/users/ahq:/bin/csh
cq:?:843:1233:Rob Till:/users/cq:/usr/local/bin/tcsh
axy:?:1272:1233:Alan Yeltsin:/users/axy:/bin/csh
helenq:?:1489:1297:Helen Schoy:/users/helenq:/bin/csh
qsc:?:1609:1375:Enid Grim:/users/qsc:/usr/local/bin/tcsh
zhq:?:2377:1318:Zong Qi:/users/zhq:/bin/csh
Broken pipe
3. Now I can visit a complicated egrep pattern, and it should make sense to you:
% egrep ‘(^[a-zA-Z]|:wi)’ /etc/printcap | head
aglw:\
:wi=AG 23:wk=multiple Apple LaserWriter IINT:
aglw1:\
:wi=AG 23:wk=Apple LaserWriter IINT:
aglw2:\
aglw3:\
aglw4:\
Broken pipe
Now you can see that the pattern specified looks either for lines that begin (^) with
an upper- or lowercase letter ([a-zA-Z]) or for lines that contain the pattern :wi.
Any time you want to look for lines that contain more than a single pattern, egrep
is the best command to use.
Task 9.6: Searching for Multiple Patterns at Once
with fgrep
Sometimes it’s helpful to look for many patterns at once. For example, you might
want to have a file of patterns and invoke a UNIX command that searches for lines
that contain any of the patterns in that file. That’s where the fgrep, or file-based grep,
command comes into play. A file of patterns can contain any pattern that grep would
understand (which means, unfortunately, that you can’t use the additional notation available
in egrep) and is specified with the -f file option.
1. I use fgrep with wrongwords, an alias and file that contains a list of words I commonly misuse. Here’s how it works:
9
177
% alias wrongwords fgrep -i -f .wrongwords
% cat .wrongwords
effect
affect
insure
ensure
idea
thought
Any time I want to check a file, for example dickens.note, to see if it has any of
these commonly misused words, I simply enter the following:
% wrongwords dickens.note
I need to determine whether these are ideas or thoughts. It’s a subtle distinction I
often forget in my writing.
2. Here’s another sample file that contains a couple words from wrongwords:
% cat sample3
At the time I was hoping to insure that the cold weather
would avoid our home, so I, perhaps foolishly, stapled the
weatherstripping along the inside of the sliding glass
door in the back room. I was surprised how much affect it
had on our enjoyment of the room, actually.
Can you see the two incorrectly used words in that sentence? The spell program can’t:
% spell sample3
The wrongwords alias, on the other hand, can detect these words:
% wrongwords sample3
At the time I was hoping to insure that the cold weather
door in the back room. I was surprised how much affect it
3. This would be a bit more useful if it could show just the individual words matched,
rather than the entire sentences. That way I wouldn’t have to figure out which
words are incorrect. To do this, I can use the awk command. It is a powerful
command that uses regular expressions, which will be discussed in greater detail in
the next chapter. This time the command will use a for loop, that is, will repeat the
command starting from the initial state (i=1) and keep adding one to the counter
(i++) until the end condition is met (i>NF): ‘{for (i=1;i<=NF;i++) print $i}’.
Each line seen by awk will be printed one word at a time with this command.
Remember that NF is the number of fields in the current line.
Here is a short example:
% echo ‘this is a sample sentence’ | awk ‘{for (i=1;i<=NF;i++) print $i}’
this
is
a
sample
sentence
9
Hour 9
178
4. I could revise my alias, but trying to get the quotation marks correct is a nightmare.
It would be much easier to make this a simple shell script instead:
% cat bin/wrongwords
# wrongwords - show a list of commonly misused words in the file
cat $* | \
awk ‘{for (i=1;i<=NF;i++) print $i}’ |\
fgrep -i -f .wrongwords
To make this work correctly, I need to remove the existing alias for wrongwords by
using the C shell unalias command, add execute permission to the shell script, and
then use rehash to ensure that the C shell can find the command when requested:
% unalias wrongwords
% chmod +x bin/wrongwords
% rehash
Now it’s ready to use:
% wrongwords sample3
insure
affect
5. The fgrep command also can exclude words from a list. If you have been using the
spell command, it’s quickly clear that the program doesn’t know anything about
acronyms or some other correctly spelled words that you might use in your writing.
That’s where fgrep can be a helpful compatriot. Build a list of words that you
commonly use that aren’t misspelled but that spell reports as being misspelled:
% alias myspell
‘spell \!* | fgrep -v -i -f $HOME/.dictionary’
% cat $HOME/.dictionary
BBS
FAX
Taylor
Utech
Zygote
Now spell can be more helpful:
% spell newsample
FAX
illetterate
Letteracy
letteracy
letterate
Papert
pre
rithmetic
Rs
Taylor
Utech
Zygote
% myspell newsample
illetterate
Letteracy
9
179
letteracy
letterate
Papert
pre
rithmetic
Rs
You have now met the entire family of grep commands. For the majority of your
searches for information, you can use the grep command itself. Sometimes, though,
it’s nice to have options, particularly if you decide to customize some of your commands as
shown in the scripts and aliases explored in this hour.
Task 9.7: Changing Things En Route with sed
I’m willing to bet that when you read about learning some UNIX programming tools
in this hour, you got anxious, your palms started to get sweaty, maybe your fingers
shook, and the little voice in your head started to say, “It’s too late! We can use a pad and
paper! We don’t need computers at all!”
Don’t panic.
If you think about it, you’ve been programming all along in UNIX. When you enter a
command to the shell, you’re programming the shell to perform immediately the task
specified. When you specify file redirection or build a pipe, you’re really writing a small
UNIX program that the shell interprets and acts upon. Frankly, when you consider how
many different commands you now know and how many different flags there are for each
of the commands, you’ve got quite a set of programming tools under your belt already, so
onward!
With the | symbol called a pipe, and commands tied together called pipelines, is it any
wonder that the information flowing down a pipeline is called a stream? For example, the
command cat test | wc means that the cat command opens the file test and streams it to
the wc program, which counts the number of lines, words, and characters therein.
To edit, or modify, the information in a pipeline, then, it seems reasonable to use a stream
editor, and that’s exactly what the sed command is! In fact, its name comes from its function:
s for stream, and ed for editor.
Here’s the bad news. The sed command is built on an old editor called ed, the same editor
that’s responsible for the grep command. Remember? The global/regular expression/
print eventually became grep. A microcosm of UNIX itself, commands to sed are separated
by a semicolon.
There are many different sed commands, but, keeping with my promise not to overwhelm
you with options and variations that aren’t going to be helpful, I’ll focus on using sed to
substitute one pattern for another and for extracting ranges of lines from a file. The general
format of the substitution command is: s/old/new/flags, where old and new are the patterns
9
Hour 9
180
you’re working with, s is the abbreviation for the substitute command, and the two most
helpful flags are g (to replace all occurrences globally on each line) and n (to tell sed to replace
only the first n occurrences of the pattern). By default, lines are listed to the screen, so a sed
expression like 10q will cause the program to list the first 10 lines and then quit (making it
an alternative to the command head -10). Deletion is similar: the command is prefaced by
one or two addresses in the file, reflecting a request to delete either all lines that match the
specified address or all in the range of the first to last.
The format of the sed command is sed, followed by the expression in quotes, then, optionally,
the name of the file to read for input.
Here are some examples.
1. I’ll start with an easy example. I’ll use grep to extract some lines from the /etc/
passwd file and then replace all colons with a single space. The format of this
command is to substitute each occurrence of : with a space, or s/:/ /:
% grep taylor /etc/passwd | sed -e ‘s/:/ /’
taylorj ?:1048:1375:James Taylor:/users/taylorj:/bin/csh
mtaylor ?:769:1375:Mary Taylor:/users/mtaylor:/usr/local/bin/tcsh
dataylor ?:375:518:Dave Taylor,,,,:/users/dataylor:/usr/local/lib/msh
taylorjr ?:203:1022:James Taylor:/users/taylorjr:/bin/csh
taylorrj ?:662:1042:Robert Taylor:/users/taylorrj:/bin/csh
taylorm ?:869:1508:Melanie Taylor:/users/taylorm:/bin/csh
taylor ?:1989:1412:Dave Taylor:/users/taylor:/bin/csh
This doesn’t quite do what I want because I neglected to append the global
instruction to the sed command to ensure that it would replace all occurrences of
the pattern on each line. I’ll try it again, this time adding a g to the instruction.
% grep taylor /etc/passwd | sed -e ‘s/:/ /g’
taylorj ? 1048 1375 James Taylor /users/taylorj /bin/csh
mtaylor ? 769 1375 Mary Taylor /users/mtaylor /usr/local/bin/tcsh
dataylor ? 375 518 Dave Taylor /users/dataylor /usr/local/lib/msh
taylorjr ? 203 1022 James Taylor /users/taylorjr /bin/csh
taylorrj ? 662 1042 Robert Taylor /users/taylorrj /bin/csh
taylorm ? 869 1508 Melanie Taylor /users/taylorm /bin/csh
taylor ? 1989 1412 Dave Taylor /users/taylor /bin/csh
2. A more sophisticated example of substitution with sed would be to modify names,
replacing all occurrences of Taylor with Tailor:
% grep taylor /etc/passwd | sed -e ‘s/Taylor/Tailor/g’
taylorj:?:1048:1375:James Tailor:/users/taylorj:/bin/csh
mtaylor:?:769:1375:Mary Tailor:/users/mtaylor:/usr/local/bin/tcsh
dataylor:?:375:518:Dave Tailor:/users/dataylor:/usr/local/lib/msh
taylorjr:?:203:1022:James Tailor:/users/taylorjr:/bin/csh
taylorrj:?:662:1042:Robert Tailor:/users/taylorrj:/bin/csh
taylorm:?:869:1508:Melanie Tailor:/users/taylorm:/bin/csh
taylor:?:1989:1412:Dave Tailor:/users/taylor:/bin/csh
9
181
The colons have returned, which is annoying, so I’ll use the fact that a semicolon
can separate multiple sed commands on the same line and try it one more time:
% grep taylor /etc/passwd | sed -e ‘s/Taylor/Tailor/g;s/:/ /g’
taylorj ? 1048 1375 James Tailor /users/taylorj /bin/csh
mtaylor ? 769 1375 Mary Tailor /users/mtaylor /usr/local/bin/tcsh
dataylor ? 375 518 Dave Tailor /users/dataylor /usr/local/lib/msh
taylorjr ? 203 1022 James Tailor /users/taylorjr /bin/csh
taylorrj ? 662 1042 Robert Tailor /users/taylorrj /bin/csh
taylorm ? 8692 1508 Melanie Tailor /users/taylorm /bin/csh
taylor ? 1989 1412 Dave Tailor /users/taylor /bin/csh
This last sed command can be read as “each time you encounter the pattern
Taylor, replace it with Tailor even if it occurs multiple times on each line. Then,
each time you encounter a colon, replace it with a space.”
3. Another example of using sed might be to rewrite the output of the who command
to be a bit more readable. Consider the results of entering who on your system:
% who
strawmye
eiyo
tzhen
kmkernek
macedot
rpm
ypchen
kodak
ttyAc
ttyAd
ttyAg
ttyAh
ttyAj
ttyAk
ttyAl
ttyAm
Nov
Nov
Nov
Nov
Nov
Nov
Nov
Nov
21
21
21
17
21
21
21
21
19:01
17:40
19:13
23:22
20:41
20:40
18:20
20:43
The output is a bit confusing; sed can help:
% who | sed
strawmye On
eiyo
On
tzhen
On
kmkernek On
macedot On
rpm
On
ypchen
On
kodak
On
‘s/tty/On Device /;s/Nov/Logged in November/’
Device Ac
Logged in November 21 19:01
Device Ad
Device Ag
Device Ah
Device Aj
Device Ak
Device Al
Device Am
This time, each occurrence of the letters tty is replaced with the phrase On Device
and, similarly, Nov is replaced with Logged in November.
4. The sed command also can be used to delete lines in the stream as it passes. The
simplest version is to specify only the command:
% who | sed ‘d’
%
There’s no output because the command matches all lines and deletes them.
Instead, to delete just the first line, simply preface the d command with that line
number:
% who | sed ‘1d’
eiyo
ttyAd
Nov 21 17:40
tzhen
ttyAg
Nov 21 19:13
kmkernek ttyAh
Nov 17 23:22
9
Hour 9
182
macedot
rpm
ypchen
kodak
ttyAj
ttyAk
ttyAl
ttyAm
Nov
Nov
Nov
Nov
21
21
21
21
20:41
20:40
18:20
20:43
To delete more than just the one line, specify the first and last lines to delete,
separating them with a comma. The following deletes the first three lines:
% who | sed ‘1,3d’
macedot ttyAj
Nov
rpm
ttyAk
Nov
ypchen
ttyAl
Nov
kodak
ttyAm
Nov
21
21
21
21
20:41
20:40
18:20
20:43
5. There’s more to deletion than that. You also can specify patterns by surrounding
them with slashes, identically to the substitution pattern. To delete the entries in
the who output between eiyo and rpm, the following would work:
% who | head -15
root
console
rick
ttyAa
brunnert ttyAb
ypchen
ttyAl
kodak
ttyAm
wh
ttyAn
klingham ttyAp
linet2
ttyAq
mdps
ttyAr
| sed ‘/eiyo/,/rpm/d’
Nov 9 07:31
Nov 21 20:58
Nov 21 20:56
Nov 21 18:20
Nov 21 20:43
Nov 21 20:33
Nov 21 19:55
Nov 21 20:17
Nov 21 20:11
You can use patterns in combination with numbers, too, so if you wanted to delete
text from the first line to the line containing kmkernek, here’s how you could do it:
% who | sed ‘1,/kmkernek/d’
macedot ttyAj
Nov 21 20:41
rpm
ttyAk
Nov 21 20:40
ypchen
ttyAl
Nov 21 18:20
kodak
ttyAm
Nov 21 20:43
6. Another aspect of sed is that the patterns are actually regular expressions. Don’t be
intimidated, though. If you understood the * and ? of filename wildcards, you’ve
learned the key basics of regular expressions: Special characters can match zero or
more letters in the pattern. Regular expressions are slightly different from shell
patterns because regular expressions are more powerful (although more confusing).
Instead of using the ? to match a character, use the . character.
Within this context, it’s rare that you need to look for patterns sufficiently complex
to require a full regular expression, which is definitely good news. The only two
characters you want to remember for regular expressions are ^, which is the
imaginary character before the first character of each line, and $, which is the
imaginary character after the end of each line.
9
JUST A MINUTE
183
Here are some pronunciation tips. UNIX folk tend to refer to the “ as
quote, the ‘ as single quote, and the ` as back quote. The * is star, the .
is dot, the ^ is caret or circumflex, the $ is dollar, and the - is dash.
What does this mean? It means that you can use sed to list everyone reported by
who that doesn’t have s as the first letter of his or her account. You can, perhaps a
bit more interestingly, eliminate all blank lines from a file with sed, too. I’ll show
you by returning to the testme file:
% cat testme
Archives/
InfoWorld/
Mail/
News/
OWL/
bin/
drop.text.hqx
keylime.pie
src/
temp/
testme
Archives/
InfoWorld/
Mail/
News/
OWL/
bin/
drop.text.hqx
keylime.pie
src/
temp/
testme
Archives/
InfoWorld/
Mail/
News/
OWL/
bin/
drop.text.hqx
keylime.pie
src/
temp/
testme
Now I’ll use sed and clean up this output:
% sed ‘/^$/d’ < testme
Archives/
InfoWorld/
Mail/
News/
Archives/
InfoWorld/
Mail/
News/
Archives/
InfoWorld/
Mail/
News/
OWL/
bin/
drop.text.hqx
OWL/
bin/
drop.text.hqx
OWL/
bin/
drop.text.hqx
keylime.pie
src/
temp/
testme
keylime.pie
src/
temp/
testme
keylime.pie
src/
temp/
testme
7. These commands can be used in combination, of course; to remove all blank lines,
all lines that contain the word keylime, and substitute BinHex for each occurrence
of hqx, one sed command can be used, albeit a complex one:
% cat testme | sed ‘/^$/d;/keylime/d;s/hqx/BinHex/g’
InfoWorld/
bin/
src/
Mail/
bitnet.mailing-lists.Z temp/
News/
drop.text.BinHex
InfoWorld/
bin/
src/
Mail/
bitnet.mailing-lists.Z temp/
testme
9
Hour 9
184
News/
InfoWorld/
Mail/
News/
drop.text.BinHex
bin/
drop.text.BinHex
testme
src/
temp/
testme
8. If you’ve ever spent any time on an electronic network, you’ve probably seen either
electronic mail or articles wherein the author responds to a previous article. Most
commonly, each line of the original message is included, each prefixed by >. It
turns out that sed is the appropriate tool either to add a prefix to a group of lines or
to remove a prefix from lines in a file.
% cat << EOF > sample
Hey Tai! I’ve been looking for a music CD and none of
the shops around here have a clue about it. I was
wondering if you’re going to have a chance to get into
Tower Records in the next week or so?
EOF
% sed ‘s/^/> /’ < sample > sample2
% cat sample2
> Hey Tai! I’ve been looking for a music CD and none of
> the shops around here have a clue about it. I was
> wondering if you’re going to have a chance to get into
> Tower Records in the next week or so?
% cat sample2 | sed ‘s/^> //’
Hey Tai! I’ve been looking for a music CD and none of
the shops around here have a clue about it. I was
wondering if you’re going to have a chance to get into
Tower Records in the next week or so?
Recall that the caret (^) signifies the beginning of the line, so the first invocation of
sed searches for the beginning of each line and replaces it with “> ”, saving the
output to the file sample2. The second use of sed—wherein I remove the prefix—
does the opposite search, finding all occurrences of “> ” that are at the beginning of
a line and replacing them with a null pattern (a null pattern is what you have when
you have two slash delimiters without anything between them).
I’ve only scratched the surface of the sed command here. It’s one of those commands
where the more you learn about it, the more powerful you realize it is. But,
paradoxically, the more you learn about it, the more you’ll really want a graphical interface
to simplify your life, too.
JUST A MINUTE
9
The only sed command I use is s (substitution). I figure that matching
patterns is best done with grep, and it’s very rare that I need to delete
specific lines from a file anyway. One helpful command I learned while
researching this portion of the hour is that sed can be used to delete from
the first line of a file to a specified pattern, meaning that it easily can be
used to strip headers from an electronic mail message by specifying the
pattern 1,/^$/d. Soon, you will learn about e-mail and how this command
can be so helpful.
185
Summary
In this hour, you really have had a chance to build on the knowledge you’re picking up about
UNIX with your introduction to two exciting and powerful UNIX utilities, grep and sed.
Finally, what’s a poker hand without some new wildcards? Because one-eyed-jacks don’t
make much sense in UNIX, you instead learned about how to specify ranges of characters in
filename patterns, further ensuring that you can type the minimum number of keys for
maximum effect.
Workshop
in the next hour.
Key Terms
exclusion set A set of characters that the pattern must not contain.
inclusion range
A range of characters that a pattern must include.
left rooted Patterns that must occur at the beginning of a line.
regular expressions A convenient notation for specifying complex patterns. Notable
special characters are ^ to match the beginning of the line and $ to match the end of the line.
wildcards Special characters that are interpreted by the UNIX shell or other programs to
have meaning other than the letter itself. For example, * is a shell wildcard and creates a
pattern that matches zero or more characters. Prefaced with a particular letter, X—X* —this
shell pattern will match all files beginning with X.
Questions
1. What wildcard expressions would you use to find the following?
■ All files in the /tmp directory
■ All files that contain a w in that directory
■ All files that start with a b, contain an e, and end with .c
■ All files that either start with test or contain the pattern hi (Notice that it
can be more than one pattern.)
2. Create regular expressions to match the following:
■ Lines that contain the words hot and cold
■ Lines that contain the word cat but not cats
■ Lines that begin with a numeral
9
Hour 9
186
3. There are two different ways you could have UNIX match all lines that contain the
words hot and cold: one uses grep and one uses pipelines. Show both.
4. Use the -v flag with various grep commands, and show the command and pattern
needed to match lines that:
■ Don’t contain cabana
■ Don’t contain either jazz or funk
■ Don’t contain jazz, funk, disco, blues, or ska.
5. Use a combination of ls -1, cat -n, and grep to find out the name of the 11th or
24th file in the /etc directory on your system.
6. There are two ways to look for lines containing any one of the words jazz, funk,
disco, blues , and ska . Show both of them.
4. What does the following do?
sed ‘s/:/ /;s/ /:/’ /etc/passwd | head
7. What does this one do?
sed ‘s/^/$ /’ < testme
In the next hour, you are introduced to some more advanced pipelining commands and the
incredibly powerful filter, awk.
9
Hour
10
187
10
Power Filters and File
Redirection
In this hour, you get to put on your programming hat and learn about two
powerful commands that can be customized infinitely and used for a wide
variety of tasks. The first of them is awk, a program that can let you grab specific
columns of information, modify text as it flows past, and even swap the order
of columns of information in a file.
The other is the tee program, which enables you to save a copy of the data being
transmitted within a pipeline.
Goals for This Hour
■ How to use the wild and weird awk command
■ How to re-route the pipeline with tee
Beginning with last hour, in which you learned the grep command, you are
learning about commands that can take months of study to master. One of the
commands treated in this hour, awk, has books written just about it, if you can
5
Hour 10
188
imagine such a thing. I say this to set the scene; this is a complex and very powerful command.
By necessity, you learn only some of the easier capabilities of these commands, but don’t
worry.
Finally, what’s a plumbing metaphor without a plumbing-related command or two? UNIX
is just the system to have odd command names. The command in question is tee.
Task 10.1: The Wild and Weird awk Command
Although the sed command can be helpful for simple editing tasks in a pipeline, for
real power, you need to invoke the awk program. The awk program is a programming
kit for analyzing and manipulating text files that have words. It’s one of the most helpful
general purpose filters in UNIX.
JUST A MINUTE
Of course, you’re wondering where awk got its name. The initial guess is
that it refers to its awkward syntax, but that’s not quite right. The name is
derived from the last names of the authors: Aho, Weinberger, and
Kernighan.
Similar to sed, awk can take its commands directly, as arguments. You also can write programs
to a file and have awk read the file for its instructions. The general approach to using the
program is awk ‘{ commands }’. There are two possible flags to awk: -f file specifies that the
instructions should be read from the file file rather than from the command line, and -Fc
indicates that the program should consider the letter c as the separator between fields of
information, rather than the default of white space (for example, one or more space or tab
characters).
1. The first awk command to learn is the most generally useful one, too, in my view:
It’s the print command. Without any arguments, it prints the lines in the file, one
by one:
% who | awk ‘{ print
root
console Nov
yuenca
ttyAo
Nov
limyx4
ttyAp
Nov
wifey
ttyAx
Nov
tobster ttyAz
Nov
taylor
ttyqh
Nov
}’
9
27
27
27
27
27
07:31
17:39
16:22
17:16
17:59
17:43
(vax1.umkc.edu)
A line of input is broken into specific fields of information, each field being
assigned a unique identifier. Field one is $1, field two $2, and so on:
10
189
% who | awk ‘{ print $1 }’
root
yuenca
limyx4
wifey
tobster
taylor
The good news is that you also can specify any other information to print by
surrounding it with double quotes:
% who | awk ‘{ print “User “ $1 “ is on terminal line “ $2 }’
User root is on terminal line console
User yuenca is on terminal line ttyAo
User limyx4 is on terminal line ttyAp
User hawk is on terminal line ttyAw
User wifey is on terminal line ttyAx
user taylor is on terminal line ttyqh
10
CAUTION
You couldn’t use single quotes to surround parameters to the print
command because they would conflict with the single quotes surrounding
the entire awk program!
2. You can see already that awk can be quite useful. Return now to the /etc/passwd
file and see how awk can help you understand the contents:
% grep taylor /etc/passwd | awk -F: ‘{ print $1 “ has “$7” as a
➥login shell.” }’
User taylorj has /bin/csh as their login shell.
User mtaylor has /usr/local/bin/tcsh as their login shell.
User dataylor has /usr/local/lib/msh as their login shell.
User taylorjr has /bin/csh as their login shell.
User taylorrj has /bin/csh as their login shell.
User taylormx has /bin/csh as their login shell.
User taylor has /bin/csh as their login shell.
3. An interesting question that came up while I was working with these examples is
how many different login shells are used at my site and which one is most popular.
On most systems, you’d be trapped, probably having to write a program to solve
this question; but with awk and some other utilities, UNIX gives you all the tools
you need:
% awk -F: ‘{print $7}’ /etc/passwd | sort | uniq -c
2
3365 /bin/csh
1 /bin/false
84 /bin/ksh
21 /bin/sh
11 /usr/local/bin/ksh
353 /usr/local/bin/tcsh
45 /usr/local/lib/msh
5
190
Hour 10
Here I’m using awk to extract just the seventh field of the password file, the home
directory, handing them all to the sort program. I then let uniq figure out which
ones occur how often and, with -c, report the count of occurrences to me. Try this
on your system, too.
4. Sticking with the password file, notice that the names therein are all in first-namethen-last-name format. That is, my account is Dave Taylor,,,,. A common
requirement that you might have is to generate a report of system users. You’d like
to sort them by name, but by last name. You do it with awk, of course:
% grep taylor /etc/passwd | awk -F: ‘{print $5}’
James Taylor,,,,
Mary Taylor,,,,
Dave Taylor,,,,
James Taylor,,,,
Robert Taylor,,,,
Melanie Taylor,,,,
Dave Taylor,,,,
That generates the list of users. Now I’ll use sed to remove those annoying commas
and awk again to reverse the order of names:
% grep taylor /etc/passwd | awk -F: ‘{print $5}’ | sed ‘s/,//g’
➥| awk ‘{print $2", “$1}’
Taylor, James
Taylor, Mary
Taylor, Dave
Taylor, James
Taylor, Robert
Taylor, Melanie
Taylor, Dave
If I feed the output of this command to sort, the names will finally be listed in the
order desired:
% grep taylor /etc/passwd | awk -F: ‘{print $5}’ | sed ‘s/,//g’
➥| awk ‘{print $2", “$1}’ | sort
Taylor, Dave
Taylor, Dave
Taylor, James
Taylor, James
Taylor, Mary
Taylor, Melanie
Taylor, Robert
This is slick. It also illustrates how you can use various UNIX commands incrementally to build up to your desired result.
5. The script earlier that looked for the login shell isn’t quite correct. It turns out that
if the user wants to have /bin/sh—the Bourne shell—as his or her default shell, the
final field can be left blank:
joe:?:45:555:Joe-Bob Billiard,,,,:/home/joe:
10
191
This can be a problem because the blank field will confuse the awk program; awk is
just counting fields in the line. The good news is that each line has an associated
number of fields, known as the NF variable. Used without a dollar sign, it indicates
how many fields are on a line; used with a dollar sign, it’s always the value of the
last field on the line itself:
% who | head -3 | awk ‘{ print NF }’
5
5
5
% who | head -3 | awk ‘{ print $NF }’
07:31
16:22
18:21
Because I’m interested in the last field in the /etc/passwd file, the best approach for
the preceding command would be to use this $NF parameter explicitly:
% grep taylor /etc/passwd | awk -F: ‘{print $NF}’ | sort | uniq -c
3365 /bin/csh
1 /bin/false
84 /bin/ksh
21 /bin/sh
10
6. Similar to NF is NR, which keeps track of the number of records (or lines) displayed.
Here’s a quick way to number a file:
% ls -l | awk ‘{ print NR”: “$0 }’
1: total 29
2: drwx------ 2 taylor
512 Nov 21 10:39 Archives/
512 Nov 16 21:55 InfoWorld/
1024 Nov 27 18:02 Mail/
512 Oct 6 09:36 News/
512 Nov 21 12:39 OWL/
512 Oct 13 10:45 bin/
8: -rw-rw---- 1 taylor
12556 Nov 16 09:49 keylime.pie
9: -rw------- 1 taylor
11503 Nov 27 18:05 randy
512 Oct 13 10:45 src/
11: drwxrwx--- 2 taylor
512 Nov 8 22:20 temp/
12: -rw-rw---- 1 taylor
0 Nov 27 18:29 testme
Here you can see that the zero field of a line is the entire line. This can be useful,
too:
% who | awk ‘{ print $2": “$0 }’
ttyAp: limyx4
ttyAp
Nov 27 16:22
ttyAt: ltbei
ttyAt
Nov 27 18:21
ttyAu: woodson ttyAu
Nov 27 18:19
ttyAv: morning ttyAv
Nov 27 18:19
ttyAw: hawk
ttyAw
Nov 27 18:12
ttyAx: wifey
ttyAx
Nov 27 17:16
ttyAz: wiwatr
ttyAz
Nov 27 18:22
ttyAA: chong
ttyAA
Nov 27 13:56
ttyAB: ishidahx ttyAB
Nov 27 18:20
5
Hour 10
192
7. Here’s another example of awk. I’ll modify the output of the ls -l command so
that I build a quick list of files and their sizes (which isn’t what is shown with the
ls -s command, recall):
% ls -lF | awk ‘{ print $9 “
rchives/ 512
InfoWorld/ 512
Mail/ 1024
News/ 512
OWL/ 512
bin/ 512
keylime.pie 12556
randy 11503
src/ 512
temp/ 512
testme 582
“ $5 }’
The output is a bit messy, so you should learn about two special character
sequences that can be embedded in the quoted arguments to print:
\n
Generates a carriage return
\t
Generates a tab character
In any case, the output is in the wrong order, anyway:
% ls -lF | awk ‘{ print $5 “\t” $9 }’
512
Archives/
512
InfoWorld/
1024
Mail/
512
News/
512
OWL/
512
bin/
12556
keylime.pie
11503
randy
512
src/
512
temp/
582
testme
Piping the preceding results to sort
largest files:
% ls -l
12556
11503
1024
582
512
-rn
could easily be used to figure out your
| awk ‘{print $5"\t” $9 }’ | sort -rn | head -5
keylime.pie
randy
Mail/
testme
temp/
8. The awk program basically looks for a pattern to appear in a line and then, if the
pattern is found, executes the instructions that follow the pattern in the awk script.
There are two special patterns in awk: BEGIN and END. The instructions that follow
BEGIN are executed before any lines of input are read. The instructions that follow
END are executed only after all the input has been read.
10
193
This can be very useful for computing the sum of a series of numbers. For example,
I’d like to know the total number of bytes I’m using for all my files:
% ls -l | awk ‘{print $5}’
512
512
1024
512
512
512
12556
11503
512
512
582
That generates the list of file sizes, but how do I sum them up? One way is to create
a new variable totalsize and output its accumulated value after each line:
% ls -l | awk ‘{ totalsize = totalsize + $5; print totalsize }’
512
1024
2048
2560
3072
3584
16140
27643
28155
28667
29249
One easy cleanup is to learn that += is a shorthand notation for “add the following
value to the variable”:
% ls -l | awk ‘{ totalsize += $5; print totalsize }’
512
1024
2048
2560
3072
3584
16140
27643
28155
28667
29249
I can use tail to get the last line only, of course, and figure out the total size
that way:
% ls -l | awk ‘{ totalsize += $5; print totalsize }’ | tail -1
29249
A better way, however, is to use the END programming block in the awk program:
% ls -l | awk ‘{ totalsize += $4 } END { print totalsize }’
29249
10
5
194
Hour 10
One more slight modification and it’s done:
% ls -l | awk ‘{ totalsize += $4 } END { print “You have a
➥total of” totalsize “ bytes used in files.” }’
You have a total of 29249 bytes used in files.
9. Here’s one further addition that can make this program even more fun:
% ls -l | awk ‘{ totalsize += $5 } END { print “You have a
➥total of” totalsize “ bytes used across “NR” files.” }’
You have a total of 29249 bytes used across 11 files.
An easier way to see all this is to create an awk program file:
% cat << EOF > script
{ totalsize += $4 }
END
{ print “You have a total of “totalsize
\
“ bytes used across “NR” files.”
}
EOF
% ls -l | awk -f script
You have a total of 29249 bytes used across 11 files.
10. Here’s one last example before I leave awk. Scripts in awk are really programs and
have all the flow-control capabilities you’d want (and then some!). One thing you
can do within an awk script is to have conditional execution of statements, the ifthen condition. The length routine returns the number of characters in the given
argument:
% awk -F: ‘{ if (length($1) == 2) print $0 }’ /etc/passwd | wc -l
26
Can you tell what this does? First off, notice that it uses the /etc/passwd file for
input and has a colon as the field delimiter (the -F:). For each line in the password
file, this awk script tests to see whether the length of the first field (the account
name) is exactly two characters long. If it is, the entire line from the password file is
printed. All lines printed are then read by the wc program, which, because I used
the -l flag, reports the total number of lines read.
What this command tells us is that on the machine, there are 26 accounts for
which the account name is two characters long.
11. The next logical question is, “How many account names have a length of each
possible number of characters?” To find out, I’ll use an advanced feature of awk just
to tantalize you: I’ll have the program build a table to keep track of the count, with
one entry per number of characters in the name:
% cat << EOF > awkscript
{
count[length($1)]++
}
END {
for (i=1; i < 9; i++)
print “There are “ count[i] “ accounts with “ i “ letter names.”
}
10
EOF
% awk
There
There
There
There
There
There
There
There
-F:
are
are
are
are
are
are
are
are
195
-f awkscript < /etc/passwd
1 accounts with 1 letter names.
You can see that longer names are preferred at this site. How about that lone
account with a single-letter account name? That’s easy to extract with the earlier
script:
% awk -F: ‘{ if (length($1) = 1) print $0 }’ < /etc/passwd
awk: syntax error near line 1
awk: illegal statement near line 1
Oops! I’ll try it again with a double equal sign:
10
% awk -F: ‘{ if (length($1) == 1) print $0 }’ < /etc/passwd
z:?:1325:1375:Chris Zed,,,,:/users/z:/bin/csh
JUST A MINUTE
The worst part of awk is its appalling error messages. Try deliberately
introducing an error into one of these awk scripts, and you’ll learn quickly
just how weird it can be! The classic error is syntax error on or near
line 1: bailing out .
The awk program is incredibly powerful. The good news is that you can easily use it
and you should find it helpful. It is a great addition to your collection of UNIX tools.
I use awk almost daily, and 99 percent of those uses are simply to extract specific columns of
information or to change the order of entries, as you saw when I reversed first name and last
name from the /etc/passwd file.
I easily could fill the rest of this book with instructions on the awk program, teaching you how
to write powerful and interesting scripts. Indeed, I could do the same with the sed program,
although I think awk has an edge in power and capabilities. The point, though, isn’t to learn
exhaustively about thousands of command options and thousands of variations, but rather
to have the key concepts and utilities at your fingertips, enabling you to build upon that
knowledge as you grow more sophisticated with UNIX.
To this goal, I note that awk is a program that has more depth and capabilities than just about
any other UNIX utility—short of actually writing programs in C. When you’ve mastered all
the lessons of this book, awk is a fruitful utility to explore further and expand your knowledge.
5
Hour 10
196
Task 10.2: Re-routing the Pipeline with tee
After the substantial sed and awk commands, this next command, tee, should be a nice
reprieve. It’s simple, can’t be programmed, and has only one possible starting flag.
Recall that the | symbol denotes a pipeline and that information traveling from one
command to another is considered to be streaming down the pipe. For example, who | sort
has the output of the who command streaming down the pipe to the sort command. Imagine
it all as some huge, albeit weird, plumbing construction.
With the plumbing metaphor in mind, you can imagine that it is helpful at times to be able
to split off the stream to make it travel down two different directions instead of just one. If
multiple pipelines really were allowed, neither you nor I ever could figure out what the heck
was going on. The simpler goal, however, of saving a copy of the stream in a file as it whizzes
past is more manageable, and that’s exactly what the tee command can do.
The only option to tee is -a, which appends the output to the specified file, rather than
replaces the contents of the file each time.
1. At its simplest, tee can grab a copy of the information being shown on the screen:
% who | tee who.out
root
console Nov
jeffhtrt ttyAo
Nov
limyx4
ttyAp
Nov
cherlbud ttyAq
Nov
garrettj ttyAr
Nov
coyote
ttyAs
Nov
ltbei
ttyAt
Nov
woodson ttyAu
Nov
morning ttyAv
Nov
wifey
ttyAx
Nov
% cat who.out
root
console Nov
jeffhtrt ttyAo
Nov
limyx4
ttyAp
Nov
cherlbud ttyAq
Nov
garrettj ttyAr
Nov
coyote
ttyAs
Nov
ltbei
ttyAt
Nov
woodson ttyAu
Nov
morning ttyAv
Nov
wifey
ttyAx
Nov
9
27
27
27
27
27
27
27
27
27
07:31
18:39
16:22
18:34
18:34
18:34
18:21
18:19
18:19
17:16
9
27
27
27
27
27
27
27
27
27
07:31
18:39
16:22
18:34
18:34
18:34
18:21
18:19
18:19
17:16
This can be quite useful for saving output.
10
197
2. Better, though, is to grab a copy of the information going down a stream in the
middle:
% ls -l
12556
8729
1024
582
512
| awk ‘{ print $5 “\t” $9 }’ | sort -rn | tee bigfiles | head -5
keylime.pie
owl.c
Mail/
tetme
temp/
This shows only the five largest files on the screen, but the bigfiles file actually has
a list of all files, sorted by size:
% cat bigfiles
12556
keylime.pie
8729
owl.c
1024
Mail/
582
tetme
512
temp/
512
src/
512
bin/
512
OWL/
512
News/
512
InfoWorld/
512
Archives/
207
sample2
199
sample
126
awkscript
10
The tee command is a classic little UNIX utility, where, as stated before, it seems
useful but a bit limited in purpose. As you’re learning through all the examples in this
book, however, from lots of little commands do big, powerful commands grow.
Summary
In this hour, you really have had a chance to build on the knowledge you’re picking up about
UNIX, with your introduction to an exciting and powerful UNIX utility, awk.
Workshop
This Workshop poses some questions about the topics presented in this chapter. It also
provides you with a preview of what you will learn in the next hour.
Questions
1. Expand on the plumbing metaphor with UNIX. What program enables you to
split the flow into multiple files? What enables you to fit multiple commands into a
pipeline? What enables you to put something into the file?
5
198
Hour 10
2. Will the following two commands do the same thing?
who | awk ‘{print $1}’ | grep taylor
who | grep taylor | awk ‘{print $1}’
3. Will this command do the same as those in the second question?
who | awk ‘{ if ($1 == “taylor”) print }’
4. Create a simple awk script that will sort lines in a file by the number of words on
the line. Pay attention to the NF record in awk itself.
Starting with the next hour, you learn about another powerful and popular program in the
entire UNIX system, a program so helpful that versions of it exist even on DOS and the
Macintosh today. It fills in the missing piece of your UNIX knowledge and, if what’s been
covered so far focuses on the plumbing analogy, this command finally moves you beyond
considering UNIX as a typewriter (a tty). What’s the program? It’s the vi screen-oriented
editor. It’s another program that deserves a book or two, but in two hours, you learn the basics
of vi and enough additional commands to let you work with the program easily and
efficiently.
10
Hour
199
11
An Introduction to the
vi Editor
If you like primitive tools, you’ve already figured out that you can use a
combination of << and cat to add lines to a file, and you can use sed and file
redirection to modify the contents of a file. These tools are rough and awkward,
and when it’s time either to create new files or to modify existing files, you need
a screen-oriented editor. In UNIX, the screen editor of choice is called vi.
There are a number of editors that may be included with your UNIX system,
including ed, ex, vi, and emacs. The latter two use the entire screen, a big
advantage, and both are powerful editors. You learn about both in these hours.
I focus on vi, however, because I believe it’s easier and, perhaps more important,
it’s guaranteed to always be part of UNIX, whereas most vendors omit emacs,
forcing you to find it yourself.
The next three hours focus on full-screen editing tools for UNIX. This is the first
of two hours in which you learn how to use vi to create and modify files. This
hour covers the basics, including how to move around in the file; how to insert
and delete characters, words, and lines; and how to search for specific patterns
11
Hour 11
200
in the text. The next hour gives an introduction to key mapping, default files, and the ways
to use the rest of UNIX while within vi. In the last hour, you learn to use an alternate UNIX
editor called emacs.
Goals for This Hour
■
■
■
■
■
■
■
■
How to start and quit vi
Simple cursor motion in vi
How to move by words and pages
How to insert text into the file
How to delete text
How to search within a file
How to have vi start out right
The key colon commands in vi
In some ways, an editor is like another operating system living within UNIX; it is so complex
that you will need two hours to learn to use vi. If you’re used to Windows or Macintosh
editors, you’ll be unhappy to find that vi doesn’t know anything about your mouse. Once
you spend some time working with vi, however, I promise it will grow on you. By the end
of this hour, you will be able to create and modify files on your UNIX system to your heart’s
content.
Task 11.1: How To Start and Quit vi
You may have noticed that many of the UNIX commands covered so far have one
characteristic in common. They all do their work, display their results, and quit.
Among the few exceptions are more and pg, where you work within the specific program
environment until you have viewed the entire contents of the file being shown or until you
quit. The vi editor is another program in this small category of environments, programs that
you move in and use until you explicitly tell the program to quit.
JUST A MINUTE
11
Where did vi get its name? It’s not quite as interesting as some of the
earlier, more colorful command names. The vi command is so named
because it’s the visual interface to the ex editor. It was written by Bill Joy
while he was at the University of California at Berkeley.
201
Before you start vi for the first time, you must learn about two aspects of its behavior. The
first is that vi is a modal editor. A mode is like an environment. Different modes in vi interpret
the same key differently. For example, if you’re in insert mode, typing a adds an a to the text,
whereas in command mode, typing a puts you in insert mode; a is the key abbreviation for the
append command. If you ever get confused about what mode you’re in, press the Escape key
on your keyboard. Pressing Escape always returns you to the command mode (and if you’re
already in command mode, it simply beeps to remind you of that fact).
When you are in command mode, you can manage your document; this includes the
capability to change text, rearrange it, and delete it. Insert mode is when you are adding text
directly into your document from the keyboard.
JUST A MINUTE
In vi, the Return key is a specific command (meaning move to the
beginning of the next line). As a result, you never need to press Return to
have vi process your command.
is a modeless editor. In emacs, the a key always adds the letter a to
the file. Commands in emacs are all indicated by holding down the
Control key while pressing the command key; for example, Control-c
deletes a character.
emacs
JUST A MINUTE
The second important characteristic of vi is that it’s a screen-oriented program. It must know
what kind of terminal, computer, or system you are using to work with UNIX. This probably
won’t be a problem for you because most systems are set up so that the default terminal type
matches the terminal or communications program you’re using. In this hour, you learn how
to recognize when vi cannot figure out what terminal you’re using and what to do about it.
You can start vi in a number of different ways, and you learn about lots of helpful alternatives
later this hour. Right now, you learn the basics. The vi command, by itself, starts the editor,
ready for you to create a new file. The vi command with a filename starts vi with the specified
file so that you can modify that file immediately.
Let’s get started!
11
Hour 11
202
1. To begin, enter vi at the prompt. If all is working well, the screen will clear, the
first character on each line will become a tilde (~), and the cursor will be sitting at
the top-left corner of the screen:
% vi
_
~
~
~
~
~
~
~
~
~
~
JUST A MINUTE
I’m going to show you only the portion of the screen that is relevant to the
command being discussed for vi, rather than show you the entire screen
each time. When the full screen is required to explain something, it’ll show
up. A smooth edge will indicate the edge of the screen, and a jagged
edge will indicate that the rest of the display has been omitted.
Type a colon character. Doing so moves the cursor to the bottom of the screen and
replaces the last tilde with the colon:
~
~
~
~
~
~
~
~
:_
11
203
Type q and press the Return key, and you should be back at the shell prompt:
~
~
~
~
~
~
~
~
:q
%
2. If that operation worked without a problem, skip to the next section, instruction 3.
If the operation did not work, you received the unknown-terminal-type error
message. You might see this on your screen:
% vi
“unknown”: Unknown terminal type
I don’t know what type of terminal you are on. All I have is “unknown”
[using open mode]
_
Alternatively, you might see this:
% vi
Visual needs addressible cursor or upline capability
:
Don’t panic. You can fix this problem. The first step is to get back to the shell
prompt. To do this, do exactly what you did in instruction 1: type :q followed by
the Return key. You should then see this:
% vi
“unknown”: Unknown terminal type
I don’t know what type of terminal you are on. All I have is “unknown”
[using open mode]
:q
%
The problem here is that vi needs to know the type of terminal you’re using, but it
can’t figure that out on its own. Therefore, you need to tell the operating system by
setting the TERM environment variable. If you know what kind of terminal you have,
use the value associated with the terminal; otherwise, try the default of vt100:
% setenv TERM vt100
If you have the $ prompt, which means you’re using the Bourne shell (sh) or Korn
shell (ksh), rather than the C shell (csh), try this:
$ TERM=vt100 ; export TERM
11
Hour 11
204
Either way, you can now try entering vi again, and it should work.
If it does work, append the command (whichever of these two commands was
successful for you) to your .login file if you use csh or to .profile if you use sh or
ksh. You can do this by entering whichever of the following commands is appropriate for your system:
% echo “setenv TERM vt100” >> .login
or
$ echo “TERM=vt100 ; export TERM” >> .profile
This way, the next time you log in, the system will remember what kind of terminal you’re using.
and other screen commands use a UNIX package called curses to
control the screen. Like most UNIX applications, curses was not designed
for a specific configuration; instead, it was designed to be deviceindependent. Therefore, to work on a specific device, you need to give it
some additional information—in this case, the terminal type.
vi
JUST A MINUTE
If vt100 didn’t work, it’s time to talk with your system administrator about
the problem or to call your UNIX vendor to find out what the specific value
should be. If you are connected through a modem or other line and you
actually are using a terminal emulator or communications package, you
might also try using ansi as a TERM setting. If that fails, call the company
that makes your software and ask the company what terminal type the
communications program is emulating.
3. Great! You have successfully launched vi, seen what it looks like, and even entered
the most important command: the quit command. Now create a simple file and
start vi so it shows you the contents of the file:
% ls -l > demo
% vi demo
total 29
drwx-----drwx-----drwx-----drwx-----drwx------rw-rw----rw-rw---drwx------rw-rw----
11
2
3
2
2
4
1
1
2
1
taylor
taylor
taylor
taylor
taylor
taylor
taylor
taylor
taylor
512
512
1024
512
512
126
165
512
0
Nov 21 10:39 Archives/
Dec 3 02:03 InfoWorld/
Dec 3 01:43 Mail/
Oct 6 09:36 News/
Dec 2 22:08 OWL/
Dec 3 16:34 awkscript
Dec 3 16:42 bigfiles
Oct 13 10:45 bin/
Dec 3 22:26 demo
-rw-rw---- 1 taylor
12556
-rw-rw---- 1 taylor
8729
-rw-rw---- 1 taylor
199
-rw-rw---- 1 taylor
207
drwx------ 2 taylor
512
drwxrwx--- 2 taylor
512
-rw-rw---- 1 taylor
582
~
~
~
~
~
~
~
“demo” 17 lines, 846 characters
Nov 16 09:49 keylime.pie
Dec 2 21:19 owl.c
Dec 3 16:11 sample
Dec 3 16:11 sample2
Oct 13 10:45 src/
Nov 8 22:20 temp/
Nov 27 18:29 tetme
You can see that vi reads the file specified on the command line. In this example,
my file is 17 lines long, but my screen can hold 25 lines. To show that some lines
lack any text, vi uses the tilde on a line by itself. Finally, note that, at the bottom,
the program shows the name of the file, the number of lines it found in the file,
and the total number of characters.
Type :q again to quit vi and return to the command line for now. When you type
the colon, the cursor will flash down to the bottom line and wait for the q as it did
before.
You have learned the most basic command in vi—the :q command—and survived
the experience. It’s all downhill from here.
Task 11.2: Simple Cursor Motion in vi
Getting to a file isn’t much good if you can’t actually move around in it. Now you
will learn how to use the cursor control keys in vi. To move left one character, type
h. To move up, type k. To move down, type j, and to move right a single character, type l
(lowercase L). You can move left one character by pressing the Backspace key, and you can
move to the beginning of the next line with the Return key.
1. Launch vi again, specifying the demo file:
% vi demo
205
11
Hour 11
206
total 29
drwx------ 2 taylor
512
drwx------ 3 taylor
512
drwx------ 2 taylor
1024
drwx------ 2 taylor
512
drwx------ 4 taylor
512
-rw-rw---- 1 taylor
126
-rw-rw---- 1 taylor
165
drwx------ 2 taylor
512
-rw-rw---- 1 taylor
0
-rw-rw---- 1 taylor
12556
-rw-rw---- 1 taylor
8729
-rw-rw---- 1 taylor
199
-rw-rw---- 1 taylor
207
drwx------ 2 taylor
512
drwxrwx--- 2 taylor
512
-rw-rw---- 1 taylor
582
~
~
~
~
~
~
~
“demo” 17 lines, 846 characters
Nov
Dec
Dec
Oct
Dec
Dec
Dec
Oct
Dec
Nov
Dec
Dec
Dec
Oct
Nov
Nov
21
3
3
6
2
3
3
13
3
16
2
3
3
13
8
27
10:39
02:03
01:43
09:36
22:08
16:34
16:42
10:45
22:26
09:49
21:19
16:11
16:11
10:45
22:20
18:29
Archives/
InfoWorld/
Mail/
News/
OWL/
awkscript
bigfiles
bin/
demo
keylime.pie
owl.c
sample
sample2
src/
temp/
tetme
You should see the cursor sitting on top of the t in total on the first line or
perhaps flashing underneath the t character. Perhaps you have a flashing-box cursor
or one that shows up in a different color. In any case, that’s your starting spot in
the file.
2. Type h once to try to move left. The cursor stays in the same spot, and vi beeps to
remind you that you can’t move left any farther on the line. Try the k key to try to
move up; the same thing will happen.
Now try typing j to move down a character:
total 29
drwx-----drwx-----drwx------
2 taylor
3 taylor
2 taylor
512 Dec 3 02:03 InfoWorld/
1024 Dec 3 01:43 Mail/
Now the cursor is on the d directory indicator of the second line of the file.
Type k to move back up to the original starting spot.
3. Using the four cursor-control keys—the h, j, k, and l keys—move around in the
file for a little bit, until you are comfortable with what’s happening on the screen.
11
207
Now try using the Backspace and Return keys to see how they help you move
around.
4. Move to the middle of a line:
total 29
2 taylor
3 taylor
2 taylor
1024 Dec 3 01:43 Mail/
Here, I’m at the middle digit in the file size of the second file in the listing. Here
are a couple of new cursor motion keys: The 0 (zero) key moves the cursor to the
beginning of the line, and $ moves it to the end of the line. First, I type 0:
total 29
2 taylor
3 taylor
2 taylor
1024 Dec 3 01:43 Mail/
Now I type $ to move to the end of the line:
total 29
2 taylor
3 taylor
2 taylor
1024 Dec 3 01:43 Mail/
5. If you have arrow keys on your keyboard, try using them to see if they work the
same way that the h, j, k, and l keys work. If the arrow keys don’t move you about,
they might have shifted you into insert mode. If you type characters and they’re
added to the file, you need to press the Escape key (or Esc, depending on your
keyboard) to return to command mode. Let’s wrap this up by leaving this edit
session. Because vi now knows that you have modified the file, it will try to ensure
that you don’t quit without saving the changes:
~
~
:q
No write since last change (:quit! overrides)
11
Hour 11
208
Use :q! (shorthand for :quit) to quit without saving the changes.
JUST A MINUTE
In general, if you try to use a colon command in vi and the program
complains that it might do something bad, try the command again,
followed by an exclamation point. I like to think of this as saying, “Do it
anyway!”
Stay in this file for the next task if you’d like, or use :q to quit.
Moving about a file using these six simple key commands is, on a small scale, much
like using the entire process of using the vi editor when working with files. Stick with
these simple commands until you’re comfortable moving around, and you will be well on
your way to becoming proficient using vi.
Task 11.3: Moving by Words and Pages
Earlier, in the description of the emacs editor, I commented that because it’s always
in insert mode, all commands must include the Control key. Well, it turns out that
vi has its share of control-key commands, commands that require you to hold down the
Control key and press another key. In this section, you learn about Ctrl-f, Ctrl-b, Ctrl-u, and
Ctrl-d. These move you forward or backward a screen and up or down half a screen of text,
respectively.
I toss a few more commands into the pot, too: w moves you forward word by word, b moves
you backward word by word, and the uppercase versions of these two commands have very
similar, but not identical, functions.
1. To see how this works, you need to create a file that is longer than the size of your
screen. An easy way to do this is to save the output of a common command to a file
over and over until the file is long enough. The system I use has lots of users, so I
needed to use the who command just once. You might have to append the output of
who to the big.output file a couple times before the file is longer than 24 lines. (You
can check using wc, of course.)
% who > big.output; wc -l big.output
40
% vi big.output
11
209
leungtc ttyrV
Dec 1 18:27
(magenta)
tuyinhwa ttyrX
Dec 3 22:38
(expert)
hollenst ttyrZ
Dec 3 22:14
(dov)
brandt
ttyrb
Nov 28 23:03
(age)
holmes
ttyrj
Dec 3 21:59
(age)
yuxi
ttyrn
Dec 1 14:19
(pc115)
frodo
ttyro
Dec 3 22:01
(mentor)
labeck
ttyrt
Dec 3 22:02
(dov)
chenlx2 ttyru
Dec 3 21:53
(mentor)
leungtc ttys0
Nov 28 15:11
(gold)
chinese ttys2
Dec 3 22:53
(excalibur)
cdemmert ttys5
Dec 3 23:00
(mentor)
yuenca
ttys6
Dec 3 23:00
(mentor)
janitor ttys7
Dec 3 18:18
(age)
mathisbp ttys8
Dec 3 23:17
(dov)
janitor ttys9
Dec 3 18:18
(age)
cs541
ttysC
Dec 2 15:16
(solaria)
yansong ttysL
Dec 1 14:44
(math)
mdps
ttysO
Nov 30 19:39
(localhost)
md
ttysU
Dec 2 08:45
(muller)
jac
ttysa
Dec 3 18:18
(localhost)
eichsted ttysb
Dec 3 23:21
(pc1)
sweett
ttysc
Dec 3 22:40
(dov)
“big.output” 40 lines, 1659 characters
11
Because I have only a 25-line display and the output is 40 lines long (you can see
that on the status line at the bottom), there is more information in this file than the
screen can display at once.
2. To see the next screenful, press Ctrl-f.
I press Control-f and get the following output:
eichsted
sweett
wellman
tuttleno
wu
daurismj
cs414
cq
cdemmert
jrlee
fitzgejs
doerrhb
cdemmert
frazierw
buckeye
mtaylor
look
ttysb
ttysc
ttysd
ttyse
ttysf
ttysg
ttysh
ttysi
ttysk
ttysn
ttyso
ttysp
ttysq
ttysr
ttyss
ttyst
ttysu
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
23:21
22:40
23:01
23:03
23:10
23:10
23:12
23:13
22:37
22:53
23:18
23:20
23:00
23:01
23:20
23:22
23:12
(pc1)
(dov)
(dov)
(indyvax)
(term01)
(dov)
(xds)
(expert)
(xsun22)
(mac1)
(dov)
(dov)
(xsun22)
(dov)
(mac2)
(dov)
(age)
Hour 11
210
janitor
ajones
~
~
~
~
ttysw
ttysx
Dec
Dec
3 18:29
3 23:23
(age)
(rass)
Notice that the entire screen was rewritten and that the familiar tilde characters
have popped up again to let me know that I have reached the end of the file.
By default, vi has a two-line overlap when you move from screen to screen. If you
look carefully, you can see that both eichsted and sweett show up on both screens
of data. This doesn’t mean that they appear twice in the file; rather, vi is trying to
help me not get lost while I wander through the file.
3. Now, what do you think will happen if I press Control-u?
leungtc
chinese
cdemmert
yuenca
janitor
mathisbp
janitor
cs541
yansong
mdps
md
jac
eichsted
sweett
wellman
tuttleno
wu
daurismj
cs414
cq
cdemmert
jrlee
fitzgejs
ttys0
ttys2
ttys5
ttys6
ttys7
ttys8
ttys9
ttysC
ttysL
ttysO
ttysU
ttysa
ttysb
ttysc
ttysd
ttyse
ttysf
ttysg
ttysh
ttysi
ttysk
ttysn
ttyso
Nov 28 15:11
Dec 3 22:53
Dec 3 23:00
Dec 3 23:00
Dec 3 18:18
Dec 3 23:17
Dec 3 18:18
Dec 2 15:16
Dec 1 14:44
Nov 30 19:39
Dec 2 08:45
Dec 3 18:18
Dec 3 23:21
Dec 3 22:40
Dec 3 23:01
Dec 3 23:03
Dec 3 23:10
Dec 3 23:10
Dec 3 23:12
Dec 3 23:13
Dec 3 22:37
Dec 3 22:53
Dec 3 23:18
(gold)
(excalibur)
(mentor)
(mentor)
(age)
(dov)
(age)
(solaria)
(math)
(localhost)
(muller)
(localhost)
(pc1)
(dov)
(dov)
(indyvax)
(term01)
(dov)
(xds)
(expert)
(xsun22)
(mac1)
(dov)
The command has moved me up half a screen. Notice where eichsted and sweett
are now. Instead of the text being replaced at once, as when I used Control-f, the
text was scrolled downward a line at a time, each new line being added as the
program went along. The Control-u command might work either way—one line or
an entire screen at a time—for you.
11
211
4. Now it’s time to try moving around in this file word by word. Type w once to see
what happens.
leungtc ttys0
chinese ttys2
cdemmert ttys5
Nov 28 15:11
Dec 3 22:53
Dec 3 23:00
(gold)
(excalibur)
(mentor)
Now type w six times more, noting that the cursor stops three times in the field to
indicate what time the user logged into the system (15:11 in this listing). Now your
cursor should be sitting on the parenthesized field:
leungtc ttys0
chinese ttys2
cdemmert ttys5
Nov 28 15:11
Dec 3 22:53
Dec 3 23:00
(gold)
(excalibur)
(mentor)
5. It’s time to move backward. Type b a few times; your cursor moves backward to
the beginning of each word.
What happens if you try to move backward and you’re already on the first word, or
if you try to move forward with the w command and you’re already on the last
word of the line? Let’s find out.
6. Using the various keys you’ve learned, move back to the beginning of the line that
starts with leungtc, which you used in instruction 4:
leungtc ttys0
chinese ttys2
cdemmert ttys5
Nov 28 15:11
Dec 3 22:53
Dec 3 23:00
(gold)
(excalibur)
(mentor)
This time, type W (uppercase W, not lowercase w) to move through this line. Can
you see the difference? Notice what happens when you hit the time field and the
parenthesized words. Instead of typing w seven times to move to the left parenthesis
before gold, you can type W only five times.
7. Try moving backward using the B command. Notice that the B command differs
from the b command the same way the W command differs from the w command.
Moving about by words, both forward and backward, being able to zip through half
screens or full screens at a time, and being able to zero in on specific spots with the
h, j, k, and l cursor-motion keys give you quite a range of motion. Practice using these
commands in various combinations to get your cursor to specific characters in your sample file.
11
Hour 11
212
Task 11.4: Inserting Text into the File
Using i, a, o, and O
Being able to move around in a file is useful. The real function of an editor, however,
is to enable you to easily add and remove—in editor parlance, insert and delete—
information. The vi editor has a special insert mode, which you must use in order to add to
the contents of the file. There are four different ways to shift into insert mode, and you learn
about all of them in this unit.
The first way to switch to insert mode is to type the letter i, which, mnemonically enough,
inserts text into the file. The other commands that accomplish more or less the same thing
are a, to append text to the file; o, to open up a line below the current line; and O, to open
up a line above the current line.
1. For this task, you need to start with a clean file, so quit from the big.output
editing session and start vi again, this time specifying a nonexistent file called
buckaroo:
% vi buckaroo
_
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
“buckaroo” [New file]
11
213
Notice that vi reminds you that this file doesn’t exist; the bottom of the screen says
New file, instead of indicating the number of lines and characters.
2. Now it’s time to try using insert mode. Try to insert a k into the file by typing k
once:
_
~
~
~
The system beeps at you because you haven’t moved into insert mode yet, and the
k still has its command meaning of moving down a line (and of course, there isn’t
another line yet).
Type i to move into insert mode, then type k again:
k_
~
~
~
There you go! You’ve added a character to the file.
3. Press the Backspace key, which will move the cursor over the letter k:
k
~
~
~
Now see what happens when you press Escape to leave insert mode and return to
the vi command mode:
_
~
~
~
Notice that the k vanished when you pressed Escape. That’s because vi only saves
text you’ve entered to the left of or above the cursor, not the letter the cursor is
resting on.
11
Hour 11
214
4. Now move back into insert mode by typing i, and enter a few sentences from a
favorite book of mine:
JUST A MINUTE
Movie buffs perhaps will recognize that the text used in this hour comes
from the book Buckaroo Banzai. The film The Adventures of Buckaroo
Banzai Across the Eighth Dimension is based on this very fun book.
“He’s not even here,” went the conservation.
“Banzai.”
“Where is he?”
“At a hotpsial in El paso.”
“What? Why werent’ we informed? What’s wrong with him?”_
~
~
I’ve deliberately left some typing errors in the text here. Fixing them will demonstrate some important features of the vi editor. If you fixed them as you went
along, that’s okay, and if you added errors of your own, that’s okay, too!
Press Escape to leave insert mode. Press Escape a second time to ensure that it
worked; remember that vi beeps to remind you that you’re already in command
mode.
5. Use the cursor motion keys (h, j, k, and l) to move the cursor to any point on the
first line:
“Banzai.”
“Where is he?”
“At the hotpsial in El paso.”
“What? Why werent’ we informed? What’s wrong with him?”
~
~
It turns out that I forgot a line of dialog between the line I’m on and the word
Banzai. One way to enter the line would be to move to the beginning of the line
“Banzai.”, insert the new text, and press Return before pressing Escape to quit
insert mode. But vi has a special command—o—to open a line immediately below
the current line for inserting text. Type o and follow along:
11
215
_
“Banzai.”
“Where is he?”
~
~
Now type the missing text:
“Who?”_
“Banzai.”
“Where is he?”
~
~
That’s it. Press Escape to return to command mode.
6. The problem with the snippet of dialog we’re using is that there’s no way to figure
out who is talking. Adding a line above this dialog helps identify the speakers.
Again, use cursor motion keys to place the cursor on the top line:
“He’s not _even here,” went the conservation.
“Banzai.”
“Where is he?”
~
~
Now you face a dilemma. You want to open up a line for new text, but you want
the line to be above the current line, not below it. It happens that vi can do that,
too. Instead of using the o command, use its big brother O instead. When I type O,
here’s what I see:
11
216
Hour 11
_
“Banzai.”
“Where is he?”
~
~
Type the new sentence and then press Escape.
I found myself stealing a peek at my own watch and overheard
General Catbird’s
aide give him the latest._
“Banzai.”
“Where is he?”
~
~
Now the dialog makes a bit more sense. The conversation, overheard by the
narrator, takes place between the general and his aide.
7. I missed a couple of words in one of the lines, so the next task is to insert them.
Use the cursor keys to move the cursor to the fifth line, just after the word Where:
General Catbird’s
aide give him the latest.
“Banzai.”
“Where_is he?”
~
~
At this juncture, I need to add the words the hell to make the sentence a bit
stronger (and correct). I can use i to insert the text, but then I end up with a
trailing space. Instead, I can add text immediately after the current cursor location
by using the a command to append, or insert, the information. When I type a, the
cursor moves one character to the right:
11
217
General Catbird’s
“Banzai.”
“Where is he?”
~
~
Here’s where vi can be difficult to use. I’m in insert mode, but there’s no way for
me to know that. When I type the letters I want to add, the screen shows that they
are appended, but what if I thought I was in insert mode when I actually was in
command mode? One trick I could use to ensure I’m in insert mode is to type the
command a second time. If the letter a shows up in the text, I simply would
backspace over it; now I would know that I’m in append mode. When I’m done
entering the new characters and I’m still in insert mode, here’s what my screen
looks like:
General Catbird’s
“Banzai.”
“Where the hell is he?”
~
~
Notice that the cursor always stayed on the i in is throughout this operation. Press
Escape to return to command mode. Notice that the cursor finally hops off the i
and moves left one character.
JUST A MINUTE
To differentiate between the i and a commands, remember that the insert
command always adds the new information immediately before the
character that the cursor is sitting upon, whereas the append command
adds the information immediately to the right of the current cursor position.
11
Hour 11
218
8. With this in mind, try to fix the apostrophe problem in the word werent’ on the
last line. Move the cursor to the n in that word:
~
To add the apostrophe immediately after the current character, do you want to use
the insert command (i) or the append (a) command? If you said “append,” give
yourself a pat on the back! Type a to append the apostrophe:
~
Type ‘ once and then press Escape.
9. Quit vi using :q, and the program reminds you that you haven’t saved your
changes to this new file:
~
~
No write since last change (:quit! overrides)
To write the changes, you need a new command, so I’ll give you a preview of a set
of colon commands you learn later in this hour. Type : (the colon character),
which moves the cursor to the bottom of the screen.
~
~
:_
Now type w to write out (save) the file, and then press the Return key:
~
~
“buckaroo” 8 lines, 272 characters
11
219
It’s okay to leave vi now. I’ll use :q to quit, and I’m safely back at the command
prompt. A quick cat confirms that the tildes were not included in the file itself:
% cat buckaroo
General Catbird’s
“Banzai.”
“What? Why weren’t’ we informed? What’s wrong with him?”
%
As you can tell, the vi editor is quite powerful, and it has a plethora of commands.
Just moving about and inserting text, you have learned 24 commands, as summarized
in Table 10.1.
Table 10.1. Summary of vi motion and insertion commands.
Command
Meaning
0
Move to the beginning of the line.
Move to the end of the line.
Append text—enter into insert mode after the current character.
Back up one screen of text.
Back up one space-delimited word.
Back up one word.
Move left one character.
Move down half a page.
Leave insert mode and return to command mode.
Move forward one screen of text.
Insert text—enter into insert mode before the current character.
Move down one line.
Move up one line.
Move right one character.
Open new line for inserting text above the current line.
Open new line for inserting text below the current line.
Move to the beginning of the next line.
$
a
^b
B
b
Backspace
^d
Escape
^f
h
i
j
k
l
O
o
Return
11
continues
Hour 11
220
Command
Meaning
û
Move up half a page.
Move forward one space-delimited word.
Move forward one word.
Write the file to disk.
Quit vi and return to the UNIX system prompt.
Quit vi and return to the UNIX system prompt, throwing away any
changes made to the file.
W
w
:w
:q
:q!
JUST A MINUTE
In this table, I use the simple shorthand notation introduced in Hour 7,
“Looking into Files.” UNIX users often use a caret followed by a character
instead of the awkward Control-c notation. Therefore, ^f has the same
meaning as Control-f. Expressing this operation as ^f does not change
the way it’s performed: you’d still press and hold down the Control key
and then type f. It’s just a shorter notation.
You’ve already learned quite a few commands, but you have barely scratched the surface of
the powerful vi command!
Task 11.5: Deleting Text
You now have many of the pieces you need to work efficiently with the vi editor, to
zip to any point in the file, and to add text wherever you’d like. Now you need to learn
how to delete characters, words, and lines.
The simplest form of the delete command is the x command, which functions as though you
are writing an X over a letter you don’t want on a printed page: It deletes the character under
the cursor. Type x five times, and you delete five characters. Deleting a line of text this way
can be quite tedious, so vi has some alternate commands. (Are you surprised?) One command
that many vi users don’t know about is the D (for “delete through the end of the line”)
command. Wherever you are on a line, if you type D, you immediately will delete everything
after the cursor to the end of that line of text.
If there’s an uppercase D command, you can just bet there’s a lowercase d command, too. The
d delete command is the first of a set of more sophisticated vi commands that you follow with
a second command that indicates what you’d like to do with the command. You already know
11
221
that w and W move you forward a word in the file; they’re known as addressing commands in
vi. You can follow d with one of these addressing commands to specify what you would like
to delete. For example, to delete a word, simply type dw.
TIME SAVER
Sometimes you might get a bit overzealous and delete more than you
anticipated. That’s not a problem—well, not too much of a problem—
because vi remembers the state of the file prior to the most recent action
taken. To undo a deletion (or insertion, for that matter), use the u command. To undo a line of changes, use the U command. Be aware that
once you’ve moved off the line in question, the U command is unable to
restore it!
1. Start vi again with the big.output file you used earlier:
leungtc ttyrV
Dec 1 18:27
(magenta)
tuyinhwa ttyrX
Dec 3 22:38
(expert)
hollenst ttyrZ
Dec 3 22:14
(dov)
brandt
ttyrb
Nov 28 23:03
(age)
holmes
ttyrj
Dec 3 21:59
(age)
yuxi
ttyrn
Dec 1 14:19
(pc)
frodo
ttyro
Dec 3 22:01
(mentor)
labeck
ttyrt
Dec 3 22:02
(dov)
chenlx2 ttyru
Dec 3 21:53
(mentor)
leungtc ttys0
Nov 28 15:11
(gold)
chinese ttys2
Dec 3 22:53
(excalibur)
cdemmert ttys5
Dec 3 23:00
(mentor)
yuenca
ttys6
Dec 3 23:00
(mentor)
janitor ttys7
Dec 3 18:18
(age)
mathisbp ttys8
Dec 3 23:17
(dov)
janitor ttys9
Dec 3 18:18
(age)
cs541
ttysC
Dec 2 15:16
(solaria)
yansong ttysL
Dec 1 14:44
(math)
mdps
ttysO
Nov 30 19:39
(localhost)
md
ttysU
Dec 2 08:45
(muller)
jac
ttysa
Dec 3 18:18
(localhost)
eichsted ttysb
Dec 3 23:21
(pc1)
sweett
ttysc
Dec 3 22:40
(dov)
11
Hour 11
222
Type x a few times to delete a few characters from the beginning of the file:
gtc ttyrV
Dec
tuyinhwa ttyrX
hollenst ttyrZ
brandt
ttyrb
holmes
ttyrj
1 18:27
(magenta)
Dec 3 22:38
(expert)
Dec 3 22:14
(dov)
Nov 28 23:03
(age)
Dec 3 21:59
(age)
Now type u to undo the last deletion:
ngtc ttyrV
Dec 1 18:27
(magenta)
tuyinhwa ttyrX
Dec 3 22:38
(expert)
hollenst ttyrZ
Dec 3 22:14
(dov)
brandt
ttyrb
Nov 28 23:03
(age)
holmes
ttyrj
Dec 3 21:59
(age)
If you type u again, what do you think will happen?
gtc ttyrV
Dec
tuyinhwa ttyrX
hollenst ttyrZ
brandt
ttyrb
holmes
ttyrj
1 18:27
(magenta)
Dec 3 22:38
(expert)
Dec 3 22:14
(dov)
Nov 28 23:03
(age)
Dec 3 21:59
(age)
The undo command alternates between the last command having happened or not
having happened. To explain it a bit better, the undo command is an action unto
itself, so the second time you type u, you’re undoing the undo command that you
just requested. Type u a few more times to convince yourself that this is the case.
2. It’s time to make some bigger changes to the file. Type dw twice to delete the
current word and the next word in the file. It should look something like this after
using the first dw:
ttyrV
Dec 1 18:27
(magenta)
tuyinhwa ttyrX
Dec 3 22:38
(expert)
hollenst ttyrZ
Dec 3 22:14
(dov)
brandt
ttyrb
Nov 28 23:03
(age)
holmes
ttyrj
Dec 3 21:59
(age)
11
223
Then it should look like this after using the second dw:
Dec 1 18:27
(magenta)
tuyinhwa ttyrX
Dec 3 22:38
hollenst ttyrZ
Dec 3 22:14
brandt
ttyrb
Nov 28 23:03
holmes
ttyrj
Dec 3 21:59
(expert)
(dov)
(age)
(age)
Type u. You see that you can undo only the most recent command. At this point,
though, because I haven’t moved from the line I’m editing, the U, or undo-a-lineof-changes, command, will restore the line to its original splendor:
leungtc
tuyinhwa
hollenst
brandt
holmes
ttyrV
ttyrX
ttyrZ
ttyrb
ttyrj
Dec 1 18:27
Dec 3 22:38
Dec 3 22:14
Nov 28 23:03
Dec 3 21:59
(magenta)
(expert)
(dov)
(age)
(age)
3. Well, in the end, I really don’t want to see some of these folks. Fortunately, I can
change the contents of this file by using the dd command to delete lines. When
using one of these two-letter commands, repeating the letter means to apply the
command to the entire line. What if I want to delete the entries for chinese and
janitor, both of which are visible on this screen?
The first step is to use the cursor keys to move down to any place on the line for
the chinese account, about halfway down the screen:
chenlx2
leungtc
chinese
cdemmert
yuenca
janitor
mathisbp
ttyru
ttys0
ttys2
ttys5
ttys6
ttys7
ttys8
Dec 3 21:53
Nov 28 15:11
Dec 3 22:53
Dec 3 23:00
Dec 3 23:00
Dec 3 18:18
Dec 3 23:17
(mentor)
(gold)
(excalibur)
(mentor)
(mentor)
(age)
(dov)
If your cursor isn’t somewhere in the middle of this line, move it so that you, too,
are not at an edge.
I had planned to remove this line completely, but perhaps I’d rather just remove
the date, time, and name of the system (in parentheses) instead. To accomplish
11
Hour 11
224
this, I don’t need to type dw a bunch of times or even x a lot of times, but rather
just D to delete through the end of the line:
chenlx2
leungtc
chinese
cdemmert
yuenca
janitor
mathisbp
ttyru
ttys0
ttys2
ttys5
ttys6
ttys7
ttys8
Dec 3 21:53
Nov 28 15:11
_
Dec 3 23:00
Dec 3 23:00
Dec 3 18:18
Dec 3 23:17
(mentor)
(gold)
(mentor)
(mentor)
(age)
(dov)
Oh, that’s not quite what I wanted to do. No problem; the undo command can fix
it. Simply typing u restores the text I deleted:
chenlx2
leungtc
chinese
cdemmert
yuenca
janitor
mathisbp
ttyru
ttys0
ttys2
ttys5
ttys6
ttys7
ttys8
Dec 3 21:53
Nov 28 15:11
Dec 3 22:53
Dec 3 23:00
Dec 3 23:00
Dec 3 18:18
Dec 3 23:17
(mentor)
(gold)
(excalibur)
(mentor)
(mentor)
(age)
(dov)
4. The problem is that I wanted to delete the two entries chinese and janitor from
the file, but I used the wrong command. Instead of using the D command, I should
use dd. Typing dd once has these results:
Dec 1 18:27
tuyinhwa ttyrX
hollenst ttyrZ
brandt
ttyrb
holmes
ttyrj
yuxi
ttyrn
frodo
ttyro
labeck
ttyrt
chenlx2 ttyru
leungtc ttys0
cdemmert ttys5
yuenca
ttys6
janitor ttys7
mathisbp ttys8
janitor ttys9
cs541
ttysC
yansong ttysL
mdps
ttysO
11
(magenta)
Dec 3 22:38
Dec 3 22:14
Nov 28 23:03
Dec 3 21:59
Dec 1 14:19
Dec 3 22:01
Dec 3 22:02
Dec 3 21:53
Nov 28 15:11
Dec 3 23:00
Dec 3 23:00
Dec 3 18:18
Dec 3 23:17
Dec 3 18:18
Dec 2 15:16
Dec 1 14:44
Nov 30 19:39
(expert)
(dov)
(age)
(age)
(pc)
(mentor)
(dov)
(mentor)
(gold)
(mentor)
(mentor)
(age)
(dov)
(age)
(solaria)
(math)
(localhost)
md
jac
eichsted
sweett
wellman
ttysU
ttysa
ttysb
ttysc
ttysd
Dec
Dec
Dec
Dec
Dec
2
3
3
3
3
08:45
18:18
23:21
22:40
23:01
225
(muller)
(localhost)
(pc1)
(dov)
(dov)
Notice that a new line of information has been pulled onto the screen at the
bottom to replace the blank line that you removed.
If you try using the u command now, what happens?
I’m almost done. A few presses of the Return key and I’m down to the entry for the
janitor account. Using dd removes that line, too:
Dec 1 18:27
tuyinhwa ttyrX
hollenst ttyrZ
brandt
ttyrb
holmes
ttyrj
yuxi
ttyrn
frodo
ttyro
labeck
ttyrt
chenlx2 ttyru
leungtc ttys0
cdemmert ttys5
yuenca
ttys6
mathisbp ttys8
janitor ttys9
cs541
ttysC
yansong ttysL
mdps
ttysO
md
ttysU
jac
ttysa
eichsted ttysb
sweett
ttysc
wellman ttysd
tuttleno ttyse
(magenta)
Dec 3 22:38
Dec 3 22:14
Nov 28 23:03
Dec 3 21:59
Dec 1 14:19
Dec 3 22:01
Dec 3 22:02
Dec 3 21:53
Nov 28 15:11
Dec 3 23:00
Dec 3 23:00
Dec 3 23:17
Dec 3 18:18
Dec 2 15:16
Dec 1 14:44
Nov 30 19:39
Dec 2 08:45
Dec 3 18:18
Dec 3 23:21
Dec 3 22:40
Dec 3 23:01
Dec 3 23:03
(expert)
(dov)
(age)
(age)
(pc)
(mentor)
(dov)
(mentor)
(gold)
(mentor)
(mentor)
(dov)
(age)
(solaria)
(math)
(localhost)
(muller)
(localhost)
(pc1)
(dov)
(dov)
(indyvax)
Each line below the one deleted moves up a line to fill in the blank space, and a
new line, for tuttleno, moves up from the following screen.
5. Now I want to return to the buckaroo file to remedy some of the horrendous
typographical errors! I don’t really care whether I save the changes I’ve just made to
this file, so I’m going to use :q! to quit, discarding these changes to the big.output
file. Entering vi buckaroo starts vi again:
11
Hour 11
226
General Catbird’s
“Banzai.”
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
There are a couple of fixes you can make in short order. The first is to change
conservation to conversation on the fourth line. To move there, press the Return
key twice, and then use W to zip forward until the cursor is at the first letter of the
word you’re editing:
General Catbird’s
“Banzai.”
Then use the dw command:
General Catbird’s
“He’s not even here,” went the .
“Banzai.”
11
227
Now enter insert mode by typing i, type the correct spelling of the word conversation,
and then press Escape:
General Catbird’s
“He’s not even here,” went the conversation.
“Banzai.”
6. That’s one fix. Now move down a couple of lines to fix the atrocious misspelling of
hospital:
“Banzai.”
~
Again, use dw to delete the word, type i to enter insert mode, type hospital, and
then press Escape. Now all is well on the line:
“Banzai.”
“At the hospital in El paso.”
~
Well, almost all is well. The first letter of Paso needs to be capitalized. Move to it
by typing w to move forward a few words:
“Banzai.”
“At the hospital in El paso.”
~
11
228
Hour 11
7. It’s time for a secret vi expert command! Instead of typing x to delete the letter, i
to enter insert mode, P as the correct letter, and then Escape to return to command
mode, there’s a much faster way to transpose case: the ~ (tilde) command. Type ~
once, and here’s what happens:
“Banzai.”
“At the hospital in El Paso.”
~
Cool, isn’t it? Back up to the beginning of the word again, using the h command,
and type ~ a few times to see what happens. Notice that each time you type ~, the
character’s case switches—transposes—and the cursor moves to the next character.
Type ~ four times, and you should end up with this:
“Banzai.”
“At the hospital in El pASO.”
~
Back up to the beginning of the word and type ~ until the word is correct.
8. One more slight change, and the file is fixed! Move to the last line of the file, to the
extra apostrophe in the word weren’t’, and type x to delete the offending character. The screen should now look like this:
General Catbird’s
“Banzai.”
“What? Why weren’t we informed? What’s wrong with him?”
~
~
~
~
~
~
~
~
~
11
229
~
~
~
~
~
~
That looks great! It’s time to save it for posterity. Use :wq, a shortcut that has vi
write out the changes and immediately quit the program:
~
~
~
%
Not only have you learned about the variety of deletion options in vi, but you also
have learned a few simple shortcut commands: ~ to transpose case and :wq to write
out the changes and quit the program all in one step.
You should feel pleased; you’re now a productive and knowledgeable vi user, and you can
modify files, making easy or tough changes. Go back to your system and experiment further,
modifying some of the other files. Be careful, though, not to make changes in any of your dot
files (for example, .cshrc), lest you cause trouble that would be difficult to fix!
Task 11.6: Searching Within a File
With the addition of two more capabilities, you’ll be ready to face down any vi
expert, demonstrating your skill and knowledge of the editor, and, much more
important, you will be able to really fly through files, moving immediately to the information
you desire.
The two new capabilities are for finding specific words or phrases in a file and for moving to
a specific line in a file. Similar to searching for patterns in more and page, the /pattern
command searches forward in the file for a specified pattern, and ?pattern searches backward
for the specified pattern. To repeat the previous search, use the n command to tell vi to search
again, in the same direction, for the next instance of the same pattern.
You can move easily to any specific line in a file, using the G, or go-to-line, command. If you
type a number before you type G, the cursor will move to that line in the file. If you type G
without a line number, the cursor will zip you to the very last line of the file (by default).
11
Hour 11
230
1. Start vi again with the big.output file:
leungtc ttyrV
Dec 1 18:27
(magenta)
tuyinhwa ttyrX
Dec 3 22:38
(expert)
hollenst ttyrZ
Dec 3 22:14
(dov)
brandt
ttyrb
Nov 28 23:03
(age)
holmes
ttyrj
Dec 3 21:59
(age)
yuxi
ttyrn
Dec 1 14:19
(pc)
frodo
ttyro
Dec 3 22:01
(mentor)
labeck
ttyrt
Dec 3 22:02
(dov)
chenlx2 ttyru
Dec 3 21:53
(mentor)
leungtc ttys0
Nov 28 15:11
(gold)
chinese ttys2
Dec 3 22:53
(excalibur)
cdemmert ttys5
Dec 3 23:00
(mentor)
yuenca
ttys6
Dec 3 23:00
(mentor)
janitor ttys7
Dec 3 18:18
(age)
mathisbp ttys8
Dec 3 23:17
(dov)
janitor ttys9
Dec 3 18:18
(age)
cs541
ttysC
Dec 2 15:16
(solaria)
yansong ttysL
Dec 1 14:44
(math)
mdps
ttysO
Nov 30 19:39
(localhost)
md
ttysU
Dec 2 08:45
(muller)
jac
ttysa
Dec 3 18:18
(localhost)
eichsted ttysb
Dec 3 23:21
(pc1)
sweett
ttysc
Dec 3 22:40
(dov)
Remember that I used :q! to quit earlier, so my changes were not retained.
To move to the very last line of the file, I type G once and see this:
cdemmert
jrlee
fitzgejs
doerrhb
cdemmert
frazierw
buckeye
mtaylor
look
janitor
ajones
~
~
~
~
11
ttysk
ttysn
ttyso
ttysp
ttysq
ttysr
ttyss
ttyst
ttysu
ttysw
ttysx
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
3
3
3
3
3
3
3
3
3
3
3
22:37
22:53
23:18
23:20
23:00
23:01
23:20
23:22
23:12
18:29
23:23
(xsun)
(mac1)
(dov)
(dov)
(xsun)
(dov)
(mac2)
(dov)
(age)
(age)
(rassilon)
231
~
~
~
~
~
~
~
~
To move to the third line of the file, I type 3 followed by G:
leungtc
tuyinhwa
hollenst
brandt
holmes
yuxi
frodo
labeck
chenlx2
leungtc
chinese
cdemmert
yuenca
janitor
mathisbp
janitor
cs541
yansong
mdps
md
jac
eichsted
sweett
ttyrV
ttyrX
ttyrZ
ttyrb
ttyrj
ttyrn
ttyro
ttyrt
ttyru
ttys0
ttys2
ttys5
ttys6
ttys7
ttys8
ttys9
ttysC
ttysL
ttysO
ttysU
ttysa
ttysb
ttysc
Dec 1 18:27
Dec 3 22:38
Dec 3 22:14
Nov 28 23:03
Dec 3 21:59
Dec 1 14:19
Dec 3 22:01
Dec 3 22:02
Dec 3 21:53
Nov 28 15:11
Dec 3 22:53
Dec 3 23:00
Dec 3 23:00
Dec 3 18:18
Dec 3 23:17
Dec 3 18:18
Dec 2 15:16
Dec 1 14:44
Nov 30 19:39
Dec 2 08:45
Dec 3 18:18
Dec 3 23:21
Dec 3 22:40
(magenta)
(expert)
(dov)
(age)
(age)
(pc)
(mentor)
(dov)
(mentor)
(gold)
(excalibur)
(mentor)
(mentor)
(age)
(dov)
(age)
(solaria)
(math)
(localhost)
(muller)
(localhost)
(pc1)
(dov)
Notice that the cursor is on the third line of the file.
2. Now it’s time to search. From my previous travels in this file, I know that the very
last line is for the account ajones, but instead of using G to move there directly, I
can search for the specified pattern by using the / search command.
Typing / immediately moves the cursor to the bottom of the screen:
md
jac
eichsted
sweett
/_
ttysU
ttysa
ttysb
ttysc
Dec
Dec
Dec
Dec
2
3
3
3
08:45
18:18
23:21
22:40
(mueller)
(localhost)
(pc1)
(dov)
11
Hour 11
232
Now I can type in the pattern ajones:
md
jac
eichsted
sweett
/ajones_
ttysU
ttysa
ttysb
ttysc
Dec
Dec
Dec
Dec
2
3
3
3
08:45
18:18
23:21
22:40
(mueller)
(localhost)
(pc1)
(dov)
When I press Return, vi spins through the file and moves me to the first line it
finds that contains the specified pattern:
cdemmert
jrlee
fitzgejs
doerrhb
cdemmert
frazierw
buckeye
mtaylor
look
janitor
ajones
~
~
~
~
~
~
~
~
~
~
~
~
ttysk
ttysn
ttyso
ttysp
ttysq
ttysr
ttyss
ttyst
ttysu
ttysw
ttysx
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
3
3
3
3
3
3
3
3
3
3
3
22:37
22:53
23:18
23:20
23:00
23:01
23:20
23:22
23:12
18:29
23:23
(xsun)
(mac1)
(dov)
(dov)
(xsun)
(dov)
(mac2)
(dov)
(age)
(age)
(rassilon)
3. If I type n to search for this pattern again, a slash appears at the very bottom line to
show that vi understood my request. But the cursor stays exactly where it is, which
indicates that this is the only occurrence of the pattern in this file.
4. Looking at this file, I noticed that the account janitor has all sorts of sessions
running. To search backward for occurrences of the account, I can use the ?
command:
~
~
?janitor_
11
233
The first search moves the cursor up one line, which leaves the screen looking
almost the same:
cdemmert
jrlee
fitzgejs
doerrhb
cdemmert
frazierw
buckeye
mtaylor
look
janitor
ajones
~
~
~
~
~
~
~
~
~
~
~
~
?janitor
ttysk
ttysn
ttyso
ttysp
ttysq
ttysr
ttyss
ttyst
ttysu
ttysw
ttysx
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
3
3
3
3
3
3
3
3
3
3
3
22:37
22:53
23:18
23:20
23:00
23:01
23:20
23:22
23:12
18:29
23:23
(xsun)
(mac1)
(dov)
(dov)
(xsun)
(dov)
(mac2)
(dov)
(age)
(age)
(rassilon)
11
Here’s where the n, or next search, can come in handy. If I type n this time and there
is another occurrence of the pattern in the file, vi moves me directly to the match:
yuxi
frodo
labeck
chenlx2
leungtc
chinese
cdemmert
yuenca
janitor
mathisbp
janitor
cs541
yansong
mdps
md
jac
eichsted
sweett
wellman
tuttleno
ttyrn
ttyro
ttyrt
ttyru
ttys0
ttys2
ttys5
ttys6
ttys7
ttys8
ttys9
ttysC
ttysL
ttysO
ttysU
ttysa
ttysb
ttysc
ttysd
ttyse
Dec 1 14:19
Dec 3 22:01
Dec 3 22:02
Dec 3 21:53
Nov 28 15:11
Dec 3 22:53
Dec 3 23:00
Dec 3 23:00
Dec 3 18:18
Dec 3 23:17
Dec 3 18:18
Dec 2 15:16
Dec 1 14:44
Nov 30 19:39
Dec 2 08:45
Dec 3 18:18
Dec 3 23:21
Dec 3 22:40
Dec 3 23:01
Dec 3 23:03
(pc)
(mentor)
(dov)
(mentor)
(gold)
(excalibur)
(mentor)
(mentor)
(age)
(dov)
(age)
(solaria)
(math)
(localhost)
(muller)
(localhost)
(pc1)
(dov)
(dov)
(indyvax)
Hour 11
234
wu
ttysf
daurismj ttysg
cs414
ttysh
Dec
Dec
Dec
3 23:10
3 23:10
3 23:12
(term01)
(dov)
(xds)
When you’re done, quit vi by using :q.
There are not dozens, but hundreds of commands in vi. Rather than overwhelm you
with all of them, even in a table, I have opted instead to work with the most basic and
important commands. By the time you’re done with this hour, your knowledge of vi
commands will be substantial, and you will be able to use the editor with little difficulty. The
next hour will expand your knowledge with more shortcuts and efficiency commands.
This task focused on searching for patterns, which is a common requirement and helpful
feature of any editor. In addition, you learned how to move to the top of the file (1G) and to
the bottom of the file (G), as well as anywhere in between.
Task 11.7: How To Start vi Correctly
The vi command wouldn’t be part of UNIX if it didn’t have some startup options
available, but there really are only two worth mentioning. The -R flag sets up vi as
a read-only file, to ensure that you don’t accidentally modify a file. The second option doesn’t
start with a dash, but with a plus sign: Any command following the plus sign is used as an
initial command to the program. This is more useful than it may sound. The command vi
+$ sample, for example, starts the editor at the bottom of the file sample, and vi +17 sample
starts the editor on the 17th line of sample.
1. First, this is the read-only format:
% vi -R buckaroo
11
General Catbird’s
“Banzai.”
~
~
~
~
~
~
235
~
~
~
~
~
~
~
~
~
“buckaroo” [Read only] 8 lines, 271 characters
Notice the addition of the [Read only] message on the status line. You can edit the
file, but if you try to save the edits with :w, you will see this:
~
~
“buckaroo” File is read only
Quit vi with :q!.
2. Next, recall that janitor occurs in many places in the big.output file. I’ll start vi
on the file line that contains the pattern janitor in the file. This time, notice where
the cursor is sitting.
% vi +/janitor big.output
brandt
ttyrb
Nov 28 23:03
(age)
holmes
ttyrj
Dec 3 21:59
(age)
yuxi
ttyrn
Dec 1 14:19
(pc)
frodo
ttyro
Dec 3 22:01
(mentor)
labeck
ttyrt
Dec 3 22:02
(dov)
chenlx2 ttyru
Dec 3 21:53
(mentor)
leungtc ttys0
Nov 28 15:11
(gold)
chinese ttys2
Dec 3 22:53
(excalibur)
cdemmert ttys5
Dec 3 23:00
(mentor)
yuenca
ttys6
Dec 3 23:00
(mentor)
janitor ttys7
Dec 3 18:18
(age)
mathisbp ttys8
Dec 3 23:17
(dov)
janitor ttys9
Dec 3 18:18
(age)
cs541
ttysC
Dec 2 15:16
(solaria)
yansong ttysL
Dec 1 14:44
(math)
mdps
ttysO
Nov 30 19:39
(localhost)
md
ttysU
Dec 2 08:45
(muller)
jac
ttysa
Dec 3 18:18
(localhost)
eichsted ttysb
Dec 3 23:21
(pc1)
sweett
ttysc
Dec 3 22:40
(dov)
wellman ttysd
Dec 3 23:01
(dov)
tuttleno ttyse
Dec 3 23:03
(indyvax)
wu
ttysf
Dec 3 23:10
(term01)
11
Hour 11
236
3. Finally, launch vi with the cursor on the third line of the file buckaroo:
% vi +3 buckaroo
General Catbird’s
“Banzai.”
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
Again, notice where the cursor rests.
At times it can be helpful to know these two starting options. In particular, I often
use +/pattern to start the editor at a specific pattern, but you can use vi for years
without ever knowing more than just the name of the command itself.
Task 11.8: The Colon Commands in vi
Without too much explanation, you have learned a couple of colon commands,
commands that have a colon as the first character. The colon immediately zooms the
cursor to the bottom of the screen for further input. These commands are actually a subset
of quite a large range of commands, all part of the ex editor on which vi is based.
The colon commands that are most helpful are as follows:
Command
Function
:e filename
Stop editing the current file and edit the specified file.
Stop editing the current file and edit the next file specified on the
command line.
:n
11
Command
Function
:q
Quit the editor.
Quit regardless of whether any changes have occurred.
Include the contents of the specified file at this position in the file
that is currently being edited.
Save the file to disk.
Save the file to disk with the specified filename.
:q!
:r filename
:w
:w filename
237
1. Start vi again, this time specifying more than one file on the command line; vi
quickly indicates that you want to edit more than one file:
% vi buckaroo big.output
2 files to edit.
Then it clears the screen and shows you the first file:
General Catbird’s
“Banzai.”
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
11
Hour 11
238
Using :w results in this:
~
~
~
2. Instead, try writing to a different file, using :w
newfile :
~
~
:w newfile_
When you press Return, you see this:
~
~
“newfile” [New file] 8 lines, 271 characters
3. Now pay attention to where the cursor is in the file. The :r, or read-file, command
always includes the contents of the file below the current line. Just before I press
Return, then, here’s what my screen looks like:
General Catbird’s
“Banzai.”
~
~
~
~
~
~
~
~
11
239
~
~
~
~
~
~
~
:r newfile_
Pressing Return yields this:
General Catbird’s
General Catbird’s
“Banzai.”
“Banzai.”
~
~
~
~
~
~
~
This can be a helpful way to include files within one another or to build a file that
contains lots of other files.
4. Now that I’ve garbled the file, I want to save it to a new file, buckaroo.confused:
~
~
:w buckaroo.confused_
11
Hour 11
240
When I press Return, I see this:
~
~
“buckaroo.confused” [New file] 16 lines, 542 characters
Older UNIX systems have a 14-character filename limit. If yours does, you
will see buckaroo.confu as the saved filename.
JUST A MINUTE
5. Now it’s time to move to the second file in the list of files given to vi at startup. To
do this, I use the :n, or next-file, command:
~
~
:n_
Pressing Return results in the next file being brought into the editor to replace
the first:
leungtc ttyrV
Dec 1 18:27
(magenta)
tuyinhwa ttyrX
Dec 3 22:38
(expert)
hollenst ttyrZ
Dec 3 22:14
(dov)
brandt
ttyrb
Nov 28 23:03
(age)
holmes
ttyrj
Dec 3 21:59
(age)
yuxi
ttyrn
Dec 1 14:19
(pc)
frodo
ttyro
Dec 3 22:01
(mentor)
labeck
ttyrt
Dec 3 22:02
(dov)
chenlx2 ttyru
Dec 3 21:53
(mentor)
leungtc ttys0
Nov 28 15:11
(gold)
chinese ttys2
Dec 3 22:53
(excalibur)
cdemmert ttys5
Dec 3 23:00
(mentor)
yuenca
ttys6
Dec 3 23:00
(mentor)
janitor ttys7
Dec 3 18:18
(age)
mathisbp ttys8
Dec 3 23:17
(dov)
janitor ttys9
Dec 3 18:18
(age)
cs541
ttysC
Dec 2 15:16
(solaria)
yansong ttysL
Dec 1 14:44
(math)
mdps
ttysO
Nov 30 19:39
(localhost)
md
ttysU
Dec 2 08:45
(muller)
jac
ttysa
Dec 3 18:18
(localhost)
eichsted ttysb
Dec 3 23:21
(pc1)
sweett
ttysc
Dec 3 22:40
(dov)
11
241
6. In the middle of working on this, I suddenly realize that I need to make a slight
change to the recently saved buckaroo.confused file. That’s where the :e command
comes in handy. Using it, I can edit any other file:
~
~
:e buckaroo.confused_
I press Return and see this:
General Catbird’s
General Catbird’s
“Banzai.”
“Banzai.”
~
~
~
~
~
~
~
“buckaroo.confused” 16 lines, 542 characters
That’s it! You now know a considerable amount about one of the most important,
and certainly most used, commands in UNIX. There’s more to learn (isn’t there
always?), but you now can edit your files with aplomb!
11
Hour 11
242
Summary
Table 10.2 summarizes the basic vi commands you learned in this hour.
Table 10.2. Basic vi commands.
Command
Meaning
0
Search forward for the next line using a specified pattern.
Search backward for the next line using a specified pattern.
Append text—enter into insert mode after the current character.
Back up one screen of text.
Back up one space-delimited word.
Back up one word.
Move down half a page.
Delete through the end of the line.
Delete—dw = delete word, dd = delete line.
Leave insert mode and return to command mode.
Move forward one screen of text.
Go to the last line of the file.
Go to the nth line of the file.
Insert text—enter into insert mode before the current character.
Move down one line.
Move up one line.
Move right one character.
Repeat last search.
Open new line for inserting text above the current line.
Open new line for inserting text below the current line.
Move to the beginning of the next line.
Move up half a page.
Undo—restore current line if changed.
$
/pattern
?pattern
a
^b
B
b
Backspace
^d
D
d
Escape
^f
G
nG
h
i
j
k
l
n
O
o
Return
û
U
11
Command
Meaning
u
Undo the last change made to the file.
Move forward one space-delimited word.
Move forward one word.
Delete a single character.
Edit a specified file without leaving vi.
Move to the next file in the file list.
Quit vi and return to the UNIX system prompt.
Quit vi and return to the UNIX system prompt, throwing away any
changes made to the file.
Include the contents of the specified file at this position in the file
that is currently being edited.
Save the file to disk with this name.
Save the file to disk.
W
w
x
:e file
:n
:q
:q!
:r file
:w file
:w
11
Workshop
in the next hour.
Key Terms
addressing commands The set of vi commands that enable you to specify what type of
object you want to work with. The d commands serve as an example: dw means delete word,
and db means delete the previous word.
colon commands The
manipulation.
243
vi
commands that begin with a colon, usually used for file
command mode The mode in which you can manage your document; this includes the
capability to change text, rearrange it, and delete it.
insert mode The vi mode that lets you enter text directly into a file. The i command starts
the insert mode, and Escape exits it.
modal A modal program has multiple environments, or modes, that offer different
capabilities. In a modal program, the Return key, for example, might do different things,
depending on which mode you are in.
Hour 11
244
modeless A modeless program always interprets a key the same way, regardless of what the
user is doing.
transpose case
Switch uppercase letters to lowercase or lowercase to uppercase.
Questions
1. What happens if you try to quit vi using :qw? Before you try it, do you expect it to
work?
2. If you’re familiar with word processing programs in the Mac or Windows environment, would you describe them as modal or modeless?
3. The d command is an example of a command that understands addressing commands. You know of quite a few. Test them to see if they will all work following d.
Make sure you see if you can figure out the command that has the opposite action
to the D command.
4. Do each of the following three commands give the same result?
D
d$
dG
5. Imagine you’re in command mode in the middle of a line that’s in the middle of
the screen. Describe what would happen if you were to type each of the following:
Badluck
Window
blad$
6. What would happen if you were to use the following startup flags?
vi +O test
vi +/joe/ names
vi +hhjjhh
vi +:q testme
The next hour expands your knowledge of the vi editor. It introduces the techniques of using
numeric repeat prefixes for commands, changing characters (rather than deleting and
inserting), searching and replacing, key mapping to enable arrow keys, and working with
UNIX while in vi.
11
Hour
245
12
Advanced vi Tricks,
Tools, and Techniques
In the previous hour, you learned some 50 vi commands that enable you to easily
move about in files, insert text, delete other text, search for specific patterns, and
move from file to file without leaving the program. This hour expands your
expertise by showing you some more powerful vi commands. Before you begin
this hour, I strongly recommend that you use vi to work with a few files. Make
sure you’re comfortable with the different modes of the program.
Goals for This Hour
■
■
■
■
Use the change and replace commands.
Use numeric repeat prefixes.
Number lines in the file.
Search and replace.
12
246
Hour 12
■ Map keys with the :map command.
■ Move sentences and paragraphs.
■ Use the :! command to access UNIX commands.
This may seem like a small list, but there’s a lot packed into it. I’ll be totally honest: You can
do fine in vi without ever reading this hour. You already know how to insert and delete text,
save or quit without saving, and you can search for particular patterns, too—even from the
command line as you start vi for the first time! On the other hand, vi is like any other complex
topic. The more you’re willing to study and learn, the more the program will bow to your
needs. This means you can accomplish a wider variety of different tasks on a daily basis.
Task 12.1: The Change and Replace Commands
In the previous hour, you saw me fix a variety of problems by deleting words and then
replacing them with new words. There is, in fact, a much smarter way to do this, and
that is by using either the change or the replace command.
Each command has a lowercase and an uppercase version, and each is quite different from
the other. The r command replaces the character that the cursor is sitting upon with the next
character you type, whereas the R command puts you into replace mode so that anything you
type overwrites whatever is already on the line until you stop typing. By contrast, C replaces
everything on the line with whatever you type. (It’s a subtle difference, but I will demonstrate
it, so don’t fear.) The c command is the most powerful of them all. The change command
c works just like the d command does, as described in the previous hour. You can use the c
command with any address command, and it will enable you to change text through to that
address, whether it’s a word, a line, or even the rest of the document.
1. Start vi with the buckaroo.confused file.
General Catbird’s
General Catbird’s
“Banzai.”
12
247
“Banzai.”
~
~
~
~
~
~
~
Without moving the cursor at all, type R. Nothing happens, or so it seems. Now
type the words Excerpt from “Buckaroo Banzai”, and watch what happens:
Excerpt from “Buckaroo Banzai”at my own watch and overheard
General Catbird’s
General Catbird’s
Now press Escape and notice that what you see on the screen is exactly what’s in
the file.
2. This isn’t, however, quite what I want. I could use either D or d$ to delete through
the end of the line, but that’s a bit awkward. Instead, I’ll use 0 to move back to the
beginning of the line. You do so, too:
Excerpt from “Buckaroo Banzai” at my own watch and overheard
General Catbird’s
General Catbird’s
This time, type C to change the contents of the line. Before you type even a single
character of the new text, notice what the line now looks like:
12
248
Hour 12
Excerpt from “Buckaroo Banzai” at my own watch and overheard
General Catbird’$
General Catbird’s
Here’s where a subtle difference comes into play! Look at the very last character on
the current line. Where the s had been, when you pressed C, the program placed a $
instead to show the range of the text to be changed by the command. Press the Tab
key once, and then type Excerpt from “Buckaroo Bansai” by Earl MacRauch.
Excerpt from “Buckaroo Bansai” by Earl MacRauchheard General
Catbird’$
General Catbird’s
This time, watch what happens when I press Escape:
Excerpt from “Buckaroo Bansai” by Earl MacRauch
General Catbird’s
3. I think I made another mistake. The actual title of the book is Buckaroo Banzai
with a z, but I’ve spelled it with an s instead. This is a chance to try the new r
command.
Use cursor control keys to move the cursor to the offending letter. I’ll use b to back
up words and then h a few times to move into the middle of the word. My screen
now looks like this:
Excerpt from “Buckaroo Bansai” by Earl MacRauch
General Catbird’s
12
249
Now type r. Again, nothing happens; the cursor doesn’t move. Type r again to
make sure it worked:
Excerpt from “Buckaroo Banrai” by Earl MacRauch
General Catbird’s
That’s no good. It replaced the s with an r, which definitely isn’t correct. Type rz,
and you should have the following:
Excerpt from “Buckaroo Banzai” by Earl MacRauch
General Catbird’s
4. Okay, those are the easy ones. Now it’s time to see what the c command can do for
you. In fact, it’s incredibly powerful. You can change just about any range of
information from the current point in the file in either direction!
To start, move to the middle of the file, where the second copy of the passage is
located:
General Catbird’s
“Banzai.”
“Banzai.”
~
~
~
~
12
250
Hour 12
~
~
~
I think I’ll just change the word aide that the cursor is sitting on to The tall beige
wall clock opted to instead. First, I type c and note that, like many other commands in vi, nothing happens. Now I type w because I want to change just the first
word. The screen should look like this:
aid$ give him the latest.
“Banzai.”
Again, the program has replaced the last character in the range of the change to a $,
so I can eyeball the situation. Now I type The tall beige wall clock opted to.
Once I reach the $, the editor stops overwriting characters and starts inserting them
instead; the screen now looks like this:
The tall beige wall clock opted to_give him the latest.
“Banzai.”
Press Escape and you’re done (though you can undo the change with the u or U
command, of course).
5. Tall and beige or not, this section makes no sense now, so change this entire line by
using the $ motion command you learned in the previous hour. First, use 0 to
move to the beginning of the line, and then type c$:
The tall beige wall clock opted to give him the latest$
“Banzai.”
12
251
This is working. The last character changed to $. Press Escape, and the entire line is
deleted:
_
“Banzai.”
6. There are still five lines below the current line. I could delete them and then type in
the information I want, but that’s primitive. Instead, the c command comes to the
rescue. Move down one line, type c5, and press Return. Watch what happens:
_
~
~
~
~
~
~
~
~
~
~
~
~
6 lines changed
In general, you always can change the current and next lines by using c followed by
a Return (because the Return key is a motion key, too, remember). By prefacing
the command with a number, I changed the range from two lines to five.
JUST A MINUTE
You might be asking, “Why two lines?” The answer is subtle. In essence,
whenever you use the c command, you change the current line plus any
additional lines that might be touched by the command. Pressing Return
moves the cursor to the following line; therefore, the current line (starting
at the cursor location) through the following line are changed. The
command probably should change just to the beginning of the following
line, but that’s beyond even my control!
12
Hour 12
252
Now press Tab four times, type (page
screen should look like this:
8) ,
and then press the Escape key. The
(page 8)
~
~
~
7. What if I change my mind? That’s where the u command comes in handy. Typing
u once undoes the last command:
General Catbird’s
“Banzai.”
“Banzai.”
~
~
~
~
~
~
~
5 more lines
The combination of replace and change commands adds a level of sophistication to
an editor that you might have suspected could only insert or delete. There’s much
more to cover in this hour, so don’t stop now!
12
253
Task 12.2: Numeric Repeat Prefixes
You have seen two commands that were prefixed by a number to cause a specific
action. The G command, in the previous hour, moves you to the very last line of the
file, unless you type in a number first. If you type in a number, the G command moves to the
specified line number. Similarly, in the previous section, you saw that typing a number before
the Return key causes vi to repeat the key the specified number of times.
Numeric repeat prefixes are actually widely available in vi and are the missing piece of your
navigational tool set.
1. I’ll move back to the top of the buckaroo.confused file. This time, I use 1G to move
there, rather than a bunch of k keys or other steps. The top of the screen now looks
like this:
General Catbird’s
Now I’ll move forward 15 words. Instead of typing w 15 times, I’ll type 15w.
General Catbird’s
2. Now I’ll move down seven lines by typing 7 and pressing Return. I’ll use o to give
myself a blank line and then press Escape:
_
“Banzai.”
12
Hour 12
254
I’d like to have Go Team Banzai! on the bottom, and I want to repeat it three times.
Can you guess how to do it? I simply type 3i to move into insert mode and then
type Go Team Banzai! . The screen looks like this:
Go Team Banzai! _
“Banzai.”
Pressing Escape has a dramatic result:
Go Team Banzai! Go Team Banzai! Go Team Banzai!
“Banzai.”
3. Now I’d like to get rid of all the lines below the current line. There are many
different ways to do this, but I’m going to try to guess how many words are present
and use a repeat count prefix to dw to delete that many words. (Actually, it’s not
critical I know the number of words, because vi will repeat the command only
while it makes sense to do so).
I type 75dw, and the screen instantly looks like this:
General Catbird’s
“Banzai.”
12
255
~
~
~
~
~
~
~
~
~
~
~
7 lines deleted
Try the undo command here to see what happens!
Almost all commands in vi can work with a numeric repeat prefix, even commands
that you might not expect to work, such as the i insert command. Remember that
a request can be accomplished in many ways. To delete five words, for example, you could
use 5dw or d5w. Experiment on your own, and you’ll get the idea.
Task 12.3: Numbering Lines in the File
It’s very helpful to have an editor that works with the entire screen, but sometimes
you need to know only what line you’re currently on. Further, sometimes it can be
very helpful to have all the lines numbered on the screen. With vi, you can do both of these—
the former by pressing ^g (remember, that’s Control-g) while in command mode, and the
latter by using a complex colon command, :set number, followed by Return. To turn off the
display of line numbers, simply type :set nonumber and press Return.
1. Much as I try to leave this file, I’m still looking at buckaroo.confused in vi. The
screen looks like this:
General Catbird’s
“Banzai.”
12
Hour 12
256
~
~
~
~
~
~
~
~
~
~
~
7 lines deleted
Can you see where the cursor is? To find out what line number the cursor is on,
press ^g, and the information is listed on the status line at the bottom:
~
~
~
“buckaroo.confused” [Modified] line 10 of 11, column 1
--90%--
There’s lots of information here, including the name of the file
(buckaroo.confused), an indication that vi thinks I’ve changed it since I started the
program ([Modified]), the current line (10), total lines in the file (11), what column
I’m in (1), and, finally, an estimate of how far into the file I am (90%).
2. Eleven lines? Count the display again. There are 12 lines. What’s going on? The
answer will become clear if I turn on line numbering for the entire file. To do this,
I type :, which zips the cursor to the bottom of the screen, where I then enter the
:set number command:
~
~
~
:set number_
12
257
Pressing Return causes the screen to change, thus:
1
2
General
3
4
5
6
7
8
9
10
11
~
~
~
~
~
~
~
~
~
~
Catbird’s
“Banzai.”
Now you can see how it figures that there are only 11 lines, even though it seems
by the screens shown in the book that there are 12 lines.
3. To turn off the line numbering, use the opposite command :set nonumber followed by Return, which restores the screen to how you’re used to seeing it.
There are definitely some times when being able to include the number of each line
is helpful. One example is if you are using awk (covered in Hour 10, “Power Filters
and File Redirection”), and it’s complaining about a specific line being in an inappropriate
format (usually by saying syntax error, bailing out!, or something similar).
Task 12.4: Search and Replace
Though most of vi is easy to learn and use, one command that always causes great
trouble for users is the search-and-replace command. The key to understanding this
command is to remember that vi is built on the line editor (ex). Instead of trying to figure
out some arcane vi command, it’s easiest to just drop to the line editor and use a simple colon
command—one identical to the command used in sed (as described in Hour 9, “Wildcards
and Regular Expressions”)—to replace an old pattern with a new one. To replace an existing
word on the current line with a new word (the simplest case), use :s/old/new/. If you want
to have all occurrences on the current line matched, you need to add the g suffix (just as with
sed): :s/old/new/g .
12
Hour 12
258
To change all occurrences of one word or phrase to another across the entire file, the
command is identical to the preceding command, except that you must prefix an indication
of the range of lines affected. Recall that $ is the last line in the file and that ranges are specified
(in this case, as in sed) by two numbers separated by a comma. It should be no surprise that
the command is :1,$ s/old/new/g.
1. You won’t be surprised to find that I’m still working with the buckaroo.confused
file, so your screen should look very similar to this:
General Catbird’s
“Banzai.”
~
~
~
~
~
~
~
~
~
~
~
~
~
~
The cursor is on the very first line. I’m going to rename Earl. I type :, the cursor
immediately moves to the bottom, and then I type s/Earl/Duke/. Pressing Return
produces this:
12
259
Excerpt from “Buckaroo Banzai” by Duke MacRauch
I found myself stealing a peek at my own watch and overheard General
Catbird’s
As you can see, this maneuver was simple and effective.
2. I’ve decided that development psychology is my bag. Now, instead of having this
Banzai character, I want my fictional character to be called Bandura. I could use
the previous command to change the occurrence on the current line, but I really
want to change all occurrences within the file.
This is no problem. I type :1,$ s/Banzai/Bandura/ and press Return. Here’s
the result:
Excerpt from “Buckaroo Bandura” by Duke MacRauch
General Catbird’s
“Bandura.”
Go Team Bandura! Go Team Banzai! Go Team Banzai!
~
~
~
~
~
~
~
~
~
~
~
~
~
The result is not quite right. Because I forgot the trailing g, vi changed only
the very first occurrence on each line, leaving the “go team” exhortation rather
confusing.
12
Hour 12
260
To try again, I type :1,$
as desired:
s/Banzai/Bandura/g ,
press Return, and the screen changes
General Catbird’s
“Bandura.”
Go Team Bandura! Go Team Bandura! Go Team Bandura!
~
~
~
~
~
~
~
~
~
~
~
7 substitutions
Notice that vi also indicates the total number of substitutions in this case.
4. I’ll press u to undo the last change.
Search and replace is one area where a windowing system like that of a Macintosh
or PC running Windows comes in handy. A windowing system offers different
boxes for the old and new patterns; it shows each change and a dialog box asking, “Should
I change this one?” Alas, this is UNIX, and it’s still designed to run on ASCII terminals.
Task 12.5: Mapping Keys with the :map Command
As you have worked through the various examples, you might have tried pressing the
arrow keys on your keyboard or perhaps a key labeled Ins or Del to insert or delete
characters. Odds are that the keys not only didn’t work, but instead caused all sorts of weird
things to happen!
The good news is that within vi is a facility that enables you to map any key to a specific action.
If these key mappings are saved in a file called .exrc in your home directory, the mappings
will be understood by vi automatically each time you use the program. The format for using
the map command is :map key command-sequence. (In a nutshell, mapping is a way of
12
261
associating an action with another action or result. For example, by plugging your computer
into the correct wall socket, you could map your action of flipping the light switch on the wall
with the result of having your computer turn on.)
JUST A MINUTE
The use of the filename .exrc is a puzzling remnant of vi having been
built on top of the ex editor. Why it couldn’t be named .virc I don’t
know.
You can also save other things in your .exrc file, including the :set number option if you’re
a nut about seeing line numbers. More interestingly, vi can be taught abbreviations so that
each time you type the abbreviation, vi expands it. The format for defining abbreviations is
:abbreviate abbreviation expanded-value. Finally, any line that begins with a double quote
is considered a comment and is ignored.
1. It’s finally time to leave the buckaroo.confused file and restart vi, this time with the
.exrc file in your home directory:
% cd
% vi .exrc
_
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
“.exrc” [New file]
12
Hour 12
262
Before I actually add any information to this new file, I’m going to define a few
abbreviations to make life a bit easier. To do this, I type :, which, as you know,
moves the cursor to the bottom of the screen. Then I’m going to define tyu as a
simple abbreviation for the lengthy phrase Teach Yourself UNIX in a Few Minutes:
~
~
~
:abbreviate tyu Teach Yourself UNIX in a Few Minutes_
Pressing Return moves the cursor back to the top.
2. Now I’ll try the abbreviation. Recall that in the .exrc, lines beginning with a
double quote are comments and are ignored when vi starts up. I press i to enter
insert mode and then type “ Sample .exrc file as shown in tyu. The screen
looks like this:
“ Sample .exrc file as shown in tyu_
~
~
As soon as I type a space or a punctuation character, or press Return, the abbreviation is expanded. In this case, I opt to move to the next line by pressing Return:
“ Sample .exrc file as shown in Teach Yourself UNIX in a Few Minutes
_
~
~
Press Escape to leave the insert mode.
3. This feature can be used also to correct common typos you make. I know that I
have a bad habit of typing teh instead of the. Because vi is smart about abbreviation expansion, I can “abbreviate” the as teh and not get into trouble:
~
~
:ab teh the_
12
263
You don’t have to type the entire word abbreviation each time. The first
two letters, ab, are sufficient for vi to figure out what’s going on!
TIME SAVER
I press Return. Now I can use my typo whenever I want, and the editor will fix it. I
can demonstrate this by adding a second comment to this file. Adding a comment
is easy because I’m still at the beginning of the second line. When I type i followed
by “ (subtly different from the example in teh, I get the following result:
“ (subtly different from the example in the_
~
~
If I type another character, instead of pressing the spacebar, vi is smart enough not
to expand the abbreviation. Try it yourself. After typing h again, I’ll see this:
“ (subtly different from the example in tehh_
~
~
Because I’m still in insert mode, however, I can backspace and replace the spare h
with a space, which instantly fixes the spelling. Finally, I type book) and press
Escape to return to command mode.
4. I have one more nifty abbreviation trick before moving to the map command. Type
:ab by itself and press Return, and vi shows you a list of the abbreviations currently
in effect:
~
~
:ab
tyu
tyu
Teach Yourself UNIX in a Few Minutes
teh
teh
the
[Hit any key to continue] _
Okay, now you can move on to key mapping.
12
Hour 12
264
5. Key mapping is as easy as defining abbreviations except that you must remember
one thing: Any control character entered must be prefaced with a ^v so that vi
doesn’t interpret it immediately. The Escape key is included in this list, too.
To map the Clear key on my keyboard to the D function, which, as you recall,
deletes text through the end of the current line, I type :map, followed by a single
space:
~
~
:map
Your keyboard might not have a Clear key. If not, please read through the
example anyway.
TIME SAVER
Now I need to type the ^v; otherwise, when I press the Clear key, it will send a
series of non-printable characters to the screen, called an escape sequence that will
confuse vi to no end. I type ^v and see this:
~
~
:map ^
The cursor is floating over the caret, which indicates that the next character typed
should be a control character. Instead of typing any specific character, however, I
simply press the Clear key. The result is that it sends the escape sequence, and vi
captures it without a problem:
~
~
:map ^[OP_
Now I type another space, because the key part of the key mapping has been
defined, and then type the command to which vi should map the Clear key:
~
~
:map ^[OP D_
12
265
Finally, I press Return, and it’s done! To test the key mapping, I’ll move back to
the very first line, to the phrase Few Minutes:
“ (subtly different from the example in the book)
~
~
To clear this line, I need only press Clear, and it works.
6. To save this as a permanent key mapping in the .exrc file, I duplicate each keystroke, but this time while in insert mode instead of at the bottom of the screen.
The result is a file that now looks like this:
“ Sample .exrc file as shown in Teach Yourself UNIX in a
“ (subtly different from the example in the book)
:map ^[OP D_
~
~
7. Mapping the arrow keys is done the same way, and, in fact, just as typing :ab and
then pressing Return shows all abbreviations. Typing :map and then Return
demonstrates that I already have my arrow keys mapped to the vi motion keys:
~
~
:map
up
^[[A
down
^[[B
left
^[[D
right
^[[C
^[OP
^[OP
[Hit any key to
12
k
j
h
l
D
continue] _
You can see that sometimes the system can be smart about defining specific keys by
name rather than by value, but the end result is the same. I now can use the arrow
keys and Clear key, and vi knows what they mean.
8. Now I’ll present one final demonstration of what you can do with keyboard
mapping. Sometimes when I’m working, I find there’s a simple, tedious activity I
must do over and over. An example might be surrounding a specific word with
quotes to meet a style guideline. This sounds more painful than it need be because
a simple key mapping can automate the entire process of quoting the current word.
Hour 12
266
I know that â isn’t used by vi, so I can map that to the new quote-a-single-word
command, making sure that I use either ^v before each control character or Escape.
I type the characters :map ^vâ i”, and I see this:
~
~
:map Â i”_
Now I again press ^v and then the Escape key. To insert a double quote, I need to
have vi go into insert mode (the i), type the quote, and then receive an Escape to
leave insert mode. The e command moves to the end of the current word, so I type
that, followed by the commands needed to append the second double quote. The
final map now looks like:
~
~
:map Â i”^[ea”^[_
Press Return and it’s done. Now move to the beginning of a word and try the new
key mapping for â.
There are a variety of customizations you can use with the vi editor, including
teaching it about special keys on your keyboard and defining task-specific keys to save
time. You can use it to abbreviate commonly used words or phrases to save time or avoid
typographical errors. Be cautious when working with the .exrc file, however, because if you
enter information that isn’t valid, it can be a bit confusing to fix it. Always try the command
directly before using it in a special key mapping, and you should stay out of trouble.
Task 12.6: Moving Sentences and Paragraphs
You have learned quite a variety of commands for moving about in files, but there
are two more vi movement commands for you to try before you learn about shell
escapes in the next unit. So far, movement has been based on screen motion, but vi hasn’t
particularly known much about the information in the file itself: Type k, and you move up
a line, regardless of what kind of file you’re viewing.
The vi editor is smarter than that, however. It has a couple of movement commands that are
defined by the text you’re currently editing. Each of these is simply a punctuation character
on your keyboard, but each is quite helpful. The first is ), which moves the cursor forward
to the beginning of the next sentence in the file. Use the opposite, (, and you can move to
the beginning of the current sentence in the file. Also worth experimenting with is }, which
moves forward a paragraph in the file, or {, which moves backwards a paragraph.
12
267
1. To try this out, create a new file that has several sentences in a paragraph and a
couple of paragraphs. Start vi and type the following text:
% cat dickens.note
Preface
story. A strong desire was upon me then, to
care and interest.
of the French people before or during the Revolution, it is truly made,
Tavistock House
November 1859
12
When I start vi on this file, here’s what my initial screen looks like:
Preface
care and interest.
Hour 12
268
“dickens.note” 28 lines, 1122 characters
Now I’ll move to the beginning of the first paragraph of text by typing /When
followed by Return. Now the screen looks like this:
Preface
2. Type ) once. The cursor moves to the beginning of the next sentence:
Try ( to move back a sentence. I end up back on the W of When starting the sentence. Repeatedly typing ( and ) should let you fly back and forth through the file,
sentence by sentence. Notice what occurs when you’re at the top few lines of the
title.
CAUTION
12
A little experimentation will demonstrate that vi defines a sentence as
anything that occurs either at the beginning of a block of text (for example, When I was...) or any word that follows a punctuation character
followed by two spaces. This two-space rule is a bit unfortunate because
modern typographic conventions have moved away from using two
spaces after the end of a sentence. If you use only one space between
sentences (as I have for this book), moving by sentence is less helpful.
269
3. I can move back to the opening word of the first paragraph by typing n to repeat
the last search pattern. The screen now looks like this:
Preface
care and interest.
To move to the next paragraph, type } once:
Preface
care and interest.
12
Hour 12
270
Type {, and you move right back to the beginning of the previous paragraph. In
fact, you can fly easily back and forth in the file by using sequences of } (or a
numeric repeat prefix like 2}) to get there faster.
These two motion commands to move by sentence and to move by paragraph are
helpful when working with stories, articles, or letters. Any time you’re working with
words rather than commands (as in the .exrc file), these commands are worth remembering.
By the way, try d) to delete a sentence, or c} to change an entire paragraph. Recall that you
always can undo the changes with u if you haven’t done anything else between the two events.
Task 12.7: Access UNIX with !
This final task on vi introduces you to one of the most powerful and least-known
commands in the editor: the ! escape-to-UNIX command. When prefaced with a
colon (:!, for example), it enables you to run UNIX commands without leaving the editor.
More powerfully, the ! command in vi itself, just like d and c, accepts address specifications,
feeds that portion of text to the command, and replaces that portion with the results of having
run that command on the text.
Let’s have a look.
1. You should still be in the dickens.intro file. I’ll start by double-checking what files
I have in my home directory. To do this, I type :!, which moves the cursor to the
bottom line:
:!_
12
I simply type ls
line:
-CF
271
and press Return, as if I were at the % prompt in the command
:!ls -CF
Archives/
big.output
dickens.note
src/
InfoWorld/
bigfiles
keylime.pie
temp/
Mail/
bin/
newfile
tetme
News/
buckaroo
owl.c
OWL/
buckaroo.confused sample
awkscript
demo
sample2
If I press Return, I’m back in the editor, and I have quickly checked what files I
have in my home directory.
2. Now for some real fun, move back to the beginning of the first paragraph and add
the text Chuck, here are my current files:. Press Return twice before using the
Escape key to return to command mode. My screen now looks like this:
Preface
Chuck, here are my current files:
_
Notice that the cursor was moved up a line. I’m now on a blank line, and the line
following is also blank.
To feed the current line to the UNIX system and replace it with the output of the
command, vi offers an easy shortcut: !!. As soon as I type the second ! (or, more
precisely, once vi figures out the desired range specified for this command), the
cursor moves to the bottom of the screen and prompts with a single ! character:
:!_
12
272
Hour 12
To list all the files in my directory, I can again type ls -CF and press Return. After
a second, vi adds the output of that command to the file:
Preface
Archives/
bigfiles
InfoWorld/
bin/
Mail/
buckaroo
News/
buckaroo.confused
OWL/
demo
awkscript
dickens.note
big.output
keylime.pie
newfile
owl.c
sample
sample2
src/
temp/
tetme
care and interest.
6 more lines
Notice that this time the status on the bottom indicates how many lines were
added to the file.
Type u to undo this change. Notice that the vi status indicator on the bottom line
says there are now six fewer lines.
3. Move back to the W in When. You are now ready to learn one of the commands that
I like most in vi. This command gives you the ability to hand a paragraph of text
to an arbitrary UNIX command.
This time, I’m going to use a sed command that was first shown in Hour 9 sed ‘s/
^/> /’, which prefaces each line with >. Ready? This is where the } command
comes in handy, too. To accomplish this trick, I type !}, moving the cursor to the
bottom of the screen, and then type the sed command as you saw earlier: sed ‘s/^/
> /’. Pressing Return feeds the lines to sed. The sed command makes the change
indicated and replaces those lines with the output of the sed command. Voilà! The
screen now looks like this:
12
273
Preface
>
>
>
>
>
>
>
>
care and interest.
!sed ‘s/^/> /’
4. I hope you’re excited to see this command in action! It’s a powerful way to interact
with UNIX while within vi.
I’ll provide a few more examples of ways to interact with UNIX while within vi.
First, I don’t really want the prefix to each line, so I’m going to type u to undo the
change.
Instead, I would rather have the system actually tighten up the lines, ensuring that
a reasonable number of words occur on each line and that no lines are too long. On
the majority of systems, there is a command called either fmt or adjust to accomplish this. To figure out which works on your system, simply use the :! command,
and feed a word or two to the fmt command to see what happens:
:!echo hi | fmt
[No write since last change]
hi
In this case, fmt did what I hoped, so I can be sure that the command exists on my
system. If your response was command unknown, adjust is a likely synonym. If
neither exists, complain to your vendor!
12
Hour 12
274
Armed with this new command, you can try another variant of !}, this time by
feeding the entire paragraph to the fmt command. I’m still at the beginning of the
word When in the text. So when I type the command !}fmt, the paragraph is cleaned
up, and the screen changes to this:
Preface
story. A strong desire was upon me then, to embody it in my own
person; and I traced out in my fancy, the state of mind of which it
would necessitate the presentation to an observant spectator, with
particular care and interest.
2 fewer lines
Again, vi tells us that the number of lines in the file has changed as a result of the
command. In this situation, tightening up the paragraph actually reduced it by two
display lines, too.
This command is so helpful that I often have it bound to a specific key with the
map command. A typical way to do this in an .exrc might be this:
:map ^P !}fmt^M
The ^M is what vi uses to record a Return. (Recall that you need to use the ^v
beforehand.) With this defined in my .exrc, I can press ^p to format the current
paragraph.
5. I will provide one more example of the ! command before I wrap up this hour.
Remember the awk command that was introduced in Hour 10? Remember how it
can easily be used to extract specific fields of information? This can be tremendously helpful in vi. Rather than continue working with the dickens.intro file,
however, I’ll quit vi and create a new file containing some output from the ls
command:
12
% ls -CF
Archives/
big.output
InfoWorld/
bigfiles
Mail/
bin/
News/
buckaroo
OWL/
buckaroo.confused
awkscript
demo
% ls -l a* b* > listing
Now I can use vi
-rw-rw----rw-rw----rw-rw----rw-rw----rw-rw---~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
~
“listing” 5
1
1
1
1
1
listing
taylor
taylor
taylor
taylor
taylor
dickens.note
keylime.pie
newfile
owl.c
sample
sample2
275
src/
temp/
tetme
to start the file with the output of the ls command:
126
1659
165
270
458
Dec
Dec
Dec
Dec
Dec
3
3
3
4
4
16:34
23:26
16:42
15:09
23:22
awkscript
big.output
bigfiles
buckaroo
buckaroo.confused
lines, 282 characters
It would be nice to use this as the basis for creating a shell script (which is just a
series of commands that you might type to the shell directly, all kept neatly in a
single file). A shell script can show me both the first and last few lines of each file,
with the middle chopped out.
The commands I’d like to have occur for each file entry are these:
echo ==== filename ====
head -5 filename; echo ...size bytes...; tail -5 filename
I’ll do this with a combination of the ! command in vi and the awk program with
the awk command:
awk ‘{ print “echo ==== “$8” ====”; print “head “$8”; echo
...”$4" bytes...; tail “$8}’
12
276
Hour 12
With the cursor on the very top line of this file, I can now type !G to pipe the
entire file through the command. The cursor drops to the bottom of the screen,
and then I type the awk script shown previously and press Return. The result is this:
echo ==== awkscript ====
head -5 awkscript; echo ...126 bytes...; tail -5 awkscript
echo ==== big.output ====
head -5 big.output; echo ...1659 bytes...; tail -5 big.output
echo ==== bigfiles ====
head -5 bigfiles; echo ...165 bytes...; tail -5 bigfiles
echo ==== buckaroo ====
head -5 buckaroo; echo ...270 bytes...; tail -5 buckaroo
echo ==== buckaroo.confused ====
head -5 buckaroo.confused; echo ...458 bytes...; tail -5 buckaroo.confused~
~
~
~
~
~
~
~
~
~
~
~
~
!awk ‘{ print “echo ==== “$8” ====”; print “head “$8”; echo
...”$4" bytes...; tail “$8}’_
If I now quit vi and ask sh to interpret the contents, here’s what happens:
% chmod +x listing
% sh listing
==== awkscript ====
{
count[length($1)]++
}
END {
for (i=1; i < 9; i++)
...126 bytes...
}
END {
for (i=1; i < 9; i++)
print “There are “ counti “ accounts with “ i “ letter names.”
}
==== big.output ====
leungtc ttyrV
Dec 1 18:27
(magenta)
tuyinhwa ttyrX
Dec 3 22:38
(expert)
hollenst ttyrZ
Dec 3 22:14
(dov)
brandt
ttyrb
Nov 28 23:03
(age)
holmes
ttyrj
Dec 3 21:59
(age)
...1659 bytes...
12
buckeye ttyss
Dec
mtaylor ttyst
Dec
look
ttysu
Dec
janitor ttysw
Dec
ajones
ttysx
Dec
==== bigfiles ====
12556
keylime.pie
8729
owl.c
1024
Mail/
582
tetme
512
temp/
...165 bytes...
512
Archives/
207
sample2
199
sample
126
awkscript
3
3
3
3
3
23:20
23:22
23:12
18:29
23:23
277
(mac2)
(dov)
(age)
(age)
(rassilon)
==== buckaroo ====
General Catbird’s
“Banzai.”
...270 bytes...
“Banzai.”
==== buckaroo.confused ====
General Catbird’s
“Bandura.”
...458 bytes...
Go Team Bandura! Go Team Bandura! Go Team Bandura!
%
Clearly the ! command opens up vi to work with the rest of the UNIX system. There’s
almost nothing that you can’t somehow manage to do within the editor, whether it’s
add or remove prefixes, clean up text, or even show what happens when you try to run a
command or reformat a passage within the current file.
12
Hour 12
278
Summary of vi Commands
A summary of the commands you learned in this hour is shown in Table 12.1.
Table 12.1. Advanced vi commands.
Command
Meaning
!!command
Replace the current line with the output of the specified UNIX
command.
Replace the current paragraph with the results of piping it through
the specified UNIX command or commands.
Move backward one sentence.
Move forward one sentence.
Change text from the point of the cursor through the end of line.
Change text in the specified range—cw changes the following word,
whereas c} changes the next paragraph.
Move to the end of the current word.
Show the current line number and other information about the file.
Replace text from the point of the cursor until Escape is pressed.
Replace the current character with the next pressed.
Prevent vi from interpreting the next character.
Move backward one paragraph.
Move forward one paragraph.
Invoke the specified UNIX command.
Define abbreviation a for phrase bcd.
Show current abbreviations, if any.
Map key a to the vi commands bcd.
Show current key mappings, if any.
Substitute new for the first instance of old on the current line.
Substitute new for all occurrences of old on the current line.
Turn off line numbering.
Turn on line numbering.
!}command
(
)
C
c
e
^g
R
r
^v
{
}
:!command
:ab a bcd
:ab
:map a bcd
:map
:s/old/new/
:s/old/new/g
:set nonumber
:set number
12
279
Summary
Clearly, vi is a very complex and sophisticated tool that enables you not only to modify your
text files, but also to customize the editor for your keyboard. Just as important, you can access
all the power of UNIX while within vi.
Workshop
in the next hour.
Key Terms
escape sequence An unprintable sequence of characters that usually specifies that your
terminal take a specific action, such as clearing the screen.
key mapping A facility that enables you to map any key to a specific action.
replace mode A mode of vi in which any characters you type replace those already in the
file.
Questions
1. What does the following command do?
:1,5 s/kitten/puppy
2. What do these commands do?
15i?ESCh
i15?ESCh
i?ESC15h
3. Try ^g on the first and last lines of a file. Explain why the percentage indicator
might not be what you expected.
4. What’s the difference between the following four strings:
rr
RrESC
cwrESC
CrESC
5. What key mappings do you have in your version of vi? Do you have labeled keys
on your keyboard that could be helpful in vi but aren’t defined? If so, define them
in your .exrc file using the :map command.
12
Hour 12
280
6. What do you think the following command will do? Try it and see if you’re right.
!}ls
With this hour and the previous one, you now know more about vi than the vast majority
of people using UNIX today. There’s a second popular editor, however, one that is modeless
and offers its own interesting possibilities for working with files and the UNIX system. It’s
called emacs, and if you have it on your system, it’s definitely worth a look. In the next hour,
you learn about this editor and some of the basics of using it.
12
Hour
281
13
An Overview of the
emacs Editor
The only screen-oriented editor that’s guaranteed to be included with the UNIX
system is vi, but that doesn’t mean that it’s the only good editor available in
UNIX! An alternative editor that has become quite popular in the last decade
(remember that UNIX is almost 25 years old) is called emacs. This hour teaches
you the fundamentals of this very different and quite powerful editing
environment.
Goals for This Hour
■
■
■
■
■
■
Launch emacs and insert text
Move around in a file
Delete characters and words
Search and replace in emacs
Use the emacs tutorial and help system
Work with other files
13
Hour 13
282
Remember what I said in the previous hour, when I introduced the emacs editor: emacs is
modeless, so be prepared for an editor that is quite unlike vi. And because it’s modeless,
there’s no insert or command mode. The result is that you have ample opportunity to use the
Control key.
JUST A MINUTE
Over the years, I have tried to become an emacs enthusiast, once even
forcing myself to use it for an entire month. I had crib sheets of commands
taped up all over my office. At the end of the month, I had attained an
editing speed that was about half of my speed in vi, an editor that I’ve
used thousands of times in the past 14 years I’ve worked in UNIX. I think
emacs has a lot going for it, and generally I think that modeless software is
better than modal software. The main obstacle I see for emacs, however, is
that it’s begging for pull-down menus like a Mac or Windows program
has. Using Control, Meta, Shift-Meta, and other weird key combinations
just isn’t as easy to use for me. On the other hand, your approach to
editing might be different, and you might not have years of vi experience
affecting your choice of editing environments. I encourage you to give
emacs a fair shake by working through all the examples I have included.
You may find that it matches your working style better than vi.
Task 13.1: Launching emacs and Inserting Text
Starting emacs is as simple as starting any other UNIX program. Simply type the
name of the program, followed by any file or files you’d like to work with. The puzzle
with emacs is figuring out what it’s actually called on your system, if you have it. There are
a couple of ways to try to identify emacs; I’ll demonstrate these in the “Action 2” section for
this task.
Once in emacs, it’s important to take a look at your computer keyboard. emacs requires you
to use not just the Control key, but another key known as the Meta key, a sort of alternative
Control key. If you have a key labeled Meta or Alt (for Alternative) on your keyboard, that’s
the one. If, like me, you don’t, simply press Escape every time a Meta key is indicated.
13
Because there are both Control and Meta keys in emacs, the notation for indicating
commands is slightly different. Throughout this book, a control-key sequence has been
shown either as Control-f or as ^f. emacs people write this differently, to allow the difference
between Control and Meta keys. In emacs notation, ^f is shown as C-f, where C- always
means Control. Similarly, M-x is the Meta key plus the character specified by x. If you don’t
have a Meta key, the sequence is Escape, followed by x. Finally, some arcane commands
involve both the Control and Meta keys being pressed (simultaneously with the other key
involved). This notation is C-M-x and indicates that you need either to press and hold down
both the Control and Meta keys while typing x or, if you don’t have a Meta (or Alt) key, to
press Escape followed by C-x.
283
With this notation in mind, you leave emacs by pressing C-x C-c (Control- x, followed by
Control-c).
1. First, see if your system has emacs available. The easiest way to find out is to type
emacs at the command line and see what happens.
% emacs
emacs: Command not found.
%
This is a good indication that emacs isn’t available. If your command worked, and
you now are in the emacs editor, move down to instruction 2 in this task.
A popular version of emacs is from the Free Software Foundation, and it’s called
gnu emacs. To see if you have this version, type gnuemacs or gnumacs at the
command line.
2. Rather than start with a blank screen, quit the program (C-x C-c) and restart emacs
with one of the earlier test files, dickens.note:
% gnuemacs dickens.note
Preface
care and interest.
----Emacs: dickens.note
(Fundamental)----Top-----------------
13
Hour 13
284
As you can see, it’s quite different from the display shown when vi starts up. The
status line at the bottom of the display offers useful information as you edit the file
at different points, and it also reminds you at all times of the name of the file, a
feature that can be surprisingly helpful. emacs can work with different kinds of files,
and here you see by the word Fundamental in the status line that emacs is prepared
for a regular text file. If you’re programming, emacs can offer special features
customized for your particular language.
3. Quit emacs by using the C-x C-c sequence, but let a few seconds pass after you
press C-x to watch what happens. When I press C-x, the bottom of the screen
suddenly changes to this:
(Fundamental)----Top----------------C-x-
Confusingly, the cursor remains at the top of the file, but emacs reminds me that
I’ve pressed C-x and that I need to enter a second command once I’ve decided what
to do. I now press C-c, and immediately exit emacs.
Already you can see there are some dramatic differences between emacs and vi. If
you’re comfortable with multiple key sequences such as C-x C-c to quit, I think you’re
going to enjoy learning emacs. If not, stick with it anyway. Even if you never use emacs, it’s
good to know a little bit about it.
JUST A MINUTE
13
Why learn about a tool you’re not going to use? In this case, the answer
is that UNIX people tend to be polarized around the question of which
editor is better. Indeed, the debate between vi and emacs is referred to as
a “religious war” because of the high levels of heat and low levels of
actual sensibility of the participants. My position is that different users will
find different tools work best for them. If emacs is closer to how you edit
files, that’s wonderful, and it’s great that UNIX offers emacs as an alternative to vi. Ultimately, the question isn’t whether one is better than the
other, but whether or not you can edit your files more quickly and easily in
one or the other.
285
Task 13.2: How To Move Around in a File
Files are composed of characters, words, lines, sentences, and paragraphs, and emacs
has commands to help you move about. Most systems have the arrow keys enabled,
which helps you avoid worrying about some of the key sequences, but it’s best to know them
all anyway.
The most basic motions are C-f and C-b, which are used to move the cursor forward and
backward one character, respectively. Switch those to the Meta command equivalents, and
the cursor will move by words: M-f moves the cursor forward a word, and M-b moves it back
a word. Pressing C-n moves the cursor to the next line, C-p to the previous line, C-a to the
beginning of the line, and C-e to the end of the line. (The vi equivalents for all of these are
l, h, w, and b for moving forward and backward a character or word; j and k for moving up
or down a line; and 0 or $ to move to the beginning or end of the current line. Which makes
more sense to you?)
To move forward a sentence, you can use M-e, which actually moves the cursor to the end
of the sentence. Pressing M-a moves it to the beginning of the sentence. Notice the parallels
between Control and Meta commands: C-a moves the cursor to the beginning of the line,
and M-a moves it to the beginning of the sentence.
Scrolling within the document is accomplished by using C-v to move forward a screen and
M-v to move back a screen. To move forward an actual page (usually 60 lines of text; this is
based on a printed page of information), you can use either C-x ] or C-x [ for forward motion
or backward motion, respectively.
Finally, to move to the very top of the file, use M-<, and to move to the bottom, use the M->
command.
1. Go back into emacs and locate the cursor. It should be at the very top of the screen:
13
_
Preface
care and interest.
286
Hour 13
(Fundamental)----Top-----------------
Move down four lines by using C-n four times. Your cursor should now be sitting
on the d of drama:
Preface
2. Next, move to the end of this sentence by using the M-e command (just like vi,
emacs expects two spaces to separate sentences):
Now type in the following text: I
fought the impulse to write this novel
vociferously, but, dear reader, I felt the injustice of the situation too
Don’t press Return or Escape when you’re done.
The screen should now look similar to this:
strongly in my breast to deny.
story. I fought the impulse to write this novel vociferously, but, dear reader,\
I felt
the injustice of the situation too strongly in my breast to deny
A strong des\
ire was upon me then, to
13
287
You can see that emacs wrapped the line when the line became too long (between
the words felt and the), and because the lines are still too long to display, a few of
them end with a backslash. The backslash isn’t actually a part of the file; with it,
emacs is telling me that those lines are longer than I might expect.
3. Now try to move back a few characters by pressing Backspace.
Uh oh! If your system is like mine, the Backspace key doesn’t move the cursor back
a character at all. Instead it starts the emacs help system, where you’re suddenly
confronted with a screen that looks like this:
You have typed C-h, the help character. Type a Help option:
A
command-apropos. Give a substring, and see a list of commands
(functions interactively callable) that contain
that substring. See also the apropos command.
B describe-bindings. Display table of all key bindings.
C describe-key-briefly. Type a command key sequence;
it prints the function name that sequence runs.
F describe-function. Type a function name and get documentation of it.
I info. The info documentation reader.
K describe-key. Type a command key sequence;
it displays the full documentation.
L view-lossage. Shows last 100 characters you typed.
M describe-mode. Print documentation of current major mode,
which describes the commands peculiar to it.
N view-emacs-news. Shows emacs news file.
S describe-syntax. Display contents of syntax table, plus explanations
T help-with-tutorial. Select the Emacs learn-by-doing tutorial.
V describe-variable. Type name of a variable;
it displays the variable’s documentation and value.
W where-is. Type command name; it prints which keystrokes
invoke that command.
--**-Emacs: *Help*
(Fundamental)----Top----------------A B C F I K L M N S T V W C-c C-d C-n C-w or Space to scroll: _
To escape the help screen (which you learn more about later in this hour), press
Escape, and your screen should be restored. Notice that the filename has been
changed and is now shown as *Help* instead of the actual file. The status line also
shows what file you’re viewing, but you aren’t always viewing the file you want to
work with.
The correct key to move the cursor back a few characters is C-b. Use that to back
up, and then use C-f to move forward again to the original cursor location.
13
Hour 13
288
4. Check that the last few lines of the file haven’t changed by using the emacs moveto-end-of-file command M->. (Think of file redirection to remember the file
motion commands). Now the screen looks like this:
Tavistock House
November 1859
_
--**-Emacs: dickens.note
(Fundamental)----Bot-----------------
5. Changing the words of Charles Dickens was fun, so save these changes and quit. If
you try to quit the program with C-x C-c, emacs reminds you that there are
unsaved changes:
(Fundamental)----Bot----------------Save file /users/taylor/dickens.note? (y or n) _
Typing y saves the changes; n quits without saving the changes; if you instead
decide to return to the edit session, Escape will cancel the action entirely. Typing n
reminds you a second time that the changes are going to be lost if you don’t save them.
(Fundamental)----Bot----------------Modified buffers exist; exit anyway? (yes or no) _
This time type y and, finally, you’re back on the command line.
13
289
Entering text in emacs is incredibly easy. It’s as if the editor is always in insert mode.
The price that you pay for this, however, is that just about anything else you do
requires Control or Meta sequences; even the Backspace key did something other than what
you wanted.
The motion commands are summarized in Table 13.1.
Table 13.1. emacs motion commands.
Command
Meaning
M->
M-<
C-v
M-v
C-x ]
C-x [
C-n
C-p
C-a
C-e
M-e
M-a
C-f
C-b
M-f
M-b
Move to the end of file.
Move to the beginning of file.
Move forward a screen.
Move backward a screen.
Move forward a page.
Move backward a page.
Move to the next line.
Move to the previous line.
Move to the end of the sentence.
Move to the beginning of the sentence.
Move forward a character.
Move backward a character.
Move forward a word.
Move backward a word.
Task 13.3: How To Delete Characters and Words
Inserting text into an emacs buffer is quite simple, and once you get the hang of it,
moving about in the file isn’t too bad, either. How about deleting text? The set of
Control and Meta commands that enable you to insert text are a precursor to all commands
in emacs, and it should come as no surprise that C-d deletes the current character, M-d deletes
the next word, M-k deletes the rest of the current sentence, and C-k deletes the rest of the
current line. If you have a key on your keyboard labeled DEL, RUBOUT, or Delete, you’re
in luck, because Delete deletes the previous character, M-Delete deletes the previous word, and
C-x Delete deletes the previous sentence.
13
Hour 13
290
Unfortunately, I have a Delete key, but it’s tied to the Backspace function on my system, so
every time I press it, it actually sends a C-h sequence to the system, not the DEL sequence.
The result is that I cannot use any of these backward-deletion commands.
JUST A MINUTE
Actually, VersaTerm Pro, the terminal emulation package I use on my
Macintosh to connect to the various UNIX systems, is smarter than that. I
can tell it whether pressing the Delete key should send a C-h or a DEL
function in the keyboard configuration screen. One flip of a toggle, and
I’m fully functional in emacs. Unfortunately, it’s not always this easy to
switch from Backspace to DEL.
1. Restart emacs with the dickens.note file, and move the cursor to the middle of the
fifth line (remember, C-n moves to the next line, and C-f moves forward a character). It should look like this:
Preface
Notice that my cursor is on the w in was on the fifth line here.
2. Press C-d C-d C-d to remove the word was. Now simply type came to revise the
sentence slightly. The screen should now look like this:
Preface
story. A strong desire came_upon me then, to
13
291
Now press Delete once to remove the last letter of the new word, and then type e
to reinsert it. Instead of backing up a character at a time, I am instead going to use
M-Delete to delete the word just added. The word is deleted, but the spaces on
either side of the word are retained.
Preface
story. A strong desire _upon me then, to
I’ll try another word to see if I can get this sentence to sound the way I’d prefer.
Type crept to see how it reads.
3. On the other hand, it’s probably not good to revise classic stories such as A Tale of
Two Cities, so the best move is for me to delete this entire sentence. If I press C-x
Delete, which is an example of a multi-keystroke command in emacs, will it do the
right thing? Recall that C-x Delete deletes the previous sentence. I press C-x
Delete, and the results are helpful, if not completely what I want to accomplish:
also requires some multistroke commands, where you might press a
control sequence and follow it with a second keystroke. Although this
allows you to have many commands to control your text, it also means you
need to know many commands.
emacs
JUST A MINUTE
Preface
story. _upon me then, to
13
292
Hour 13
That’s okay. Now I can delete the second part of the sentence by using the M-k
command. Now the screen looks like what I want:
story. _
4. Here’s a great feature of emacs! I just realized that deleting sentences is just as wildly
inappropriate as changing words, so I want to undo the last two changes. If I were
using vi, I’d be stuck, because vi remembers only the last change; but emacs has
that beat. With emacs, you can back up as many changes as you’d like, usually until
you restore the original file. To step backwards, use C-x u.
The first time I press C-x u, the screen changes to this:
story. _upon me then, to
care and interest.
The second time I press it, the screen goes back even further in my revision history:
story. A strong desire crept_upon me then, to
care and interest.
13
Finally, using C-x
u
293
three more times causes the original text to be restored:
Preface
care and interest.
(Fundamental)----Top----------------Undo!
If you don’t have a Delete key, some of the deletion commands will be unavailable
to you, regrettably. Generally, though, emacs has as many ways to delete text as vi has,
if not more. The best feature, however, is that, unlike vi, emacs remembers edit changes from
the beginning of your editing session. You can always back up as far as you want by using the
C-x u undo request.
The deletion command keys are summarized in Table 13.2.
Table 13.2. Deletion commands in emacs.
Command
Meaning
Delete
C-d
M-Delete
M-d
C-x Delete
M-k
C-k
C-x u
Delete the previous character.
Delete the current character.
Delete the previous word.
Delete the next word.
Delete the previous sentence.
Delete the rest of the current sentence.
Delete the rest of the current line.
Undo the last edit change.
13
Hour 13
294
Task 13.4: Search and Replace in emacs
Because emacs reserves the last line of the screen for its own system prompts,
searching and replacing is easier than in vi. Moreover, the system prompts for the
fields and asks, for each occurrence, whether to change it. On the other hand, this command
isn’t a simple key press or two, but rather it is an example of a named emacs command. A
named emacs command is a command that requires you to type its name, such as queryreplace, rather than a command key or two.
Searching forward for a pattern is done by pressing C-s, and searching backward is done with
C-r (the mnemonics are search forward or reverse search). To leave the search once you’ve
found what you want, press Escape, and to cancel the search, returning to your starting point,
use C-g.
CAUTION
Unfortunately, you might find that pressing C-s does very strange things to
your system. In fact, ^s and ^q often are used as flow control on a
terminal, and by pressing the C-s key, you’re actually telling the terminal
emulator to stop sending information until it sees a C-q. Flow control is the
protocol used by your computer and terminal to make sure that neither
outpaces the other during data transmission. If this happens to you, you
need to try to turn off XON/XOFF flow control. Ask your system administrator for help.
Query and replace is really a whole new feature within emacs. To start a query and replace,
use M-x query-replace. emacs will prompt for what to do next. Once a match is shown, you
can type a variety of different commands to affect what happens: y makes the change; n means
to leave it as is but move to the next match; Escape or q quits replace mode; and !
automatically replaces all further occurrences of the pattern without further prompting.
1. I’m still looking at the dickens.note file, and I have moved the cursor to the topleft corner by using the M-<. Somewhere in the file is the word Revolution, but I’m
not sure where. Worse, every time I press C-s, the terminal freezes up until I press
C-q, because of flow control problems. Instead of searching forward, I’ll search
backward by first moving the cursor to the bottom of the file with M-> and then
pressing C-r.
I-search backward:
13
(Fundamental)----Bot-----------------
295
As I type each character of the pattern Revolution, the cursor dances backward,
matching the pattern as it grows longer and longer, until emacs finds the word I seek:
Tavistock House
November 1859
I-search backward: Revol
(Fundamental)----Bot-----------------
2. Now try the query-replace feature. To begin, I move to the top of the file with
M-< and then press M-x, which causes the notation to show up on the bottom
status line:
(Fundamental)----Top----------------M-x _
I then type the words query-replace and press Return. emacs understands that I
want to find all occurrences of a pattern and replace them with another. emacs
changes the prompt to this:
(Fundamental)----Top----------------Query replace: _
13
Hour 13
296
Now I type in the word that I want to replace. To cause confusion in the file, I
think I’ll change French to Danish, because maybe A Tale of Two Cities really takes
place in London and Copenhagen! To do this, I type French and press Return. The
prompt again changes to this:
(Fundamental)----Top----------------Query replace French with: _
I type Danish, and again press Return.
of the French_people before or during the Revolution, it is truly made,
(Fundamental)----Top----------------Query replacing French with Danish:
It may not be completely obvious, but emacs has found a match (immediately
before the cursor) and is prompting me for what to do next. The choices here are
summarized in Table 13.3.
Table 13.3. Options during query and replace.
Command
Meaning
y
Change this occurrence of the pattern.
Don’t change this occurrence, but look for another.
Don’t change this occurrence. Leave query-replace completely (you
also can use Escape for this function).
Change this occurrence and all others in the file.
n
q
!
I opt to make this and all other possible changes in the file, by pressing !, and the
screen changes to tell me that there were no more occurrences:
13
297
of the Danish_people before or during the Revolution, it is truly made,
(Fundamental)----Top---------------Done
Searching in emacs is awkward, particularly due to the flow control problems that you
may incur because of your terminal. However, searching and replacing with the
query-replace command is fantastic—much better and more powerful than the vi alternative. As I said earlier, your assessment of emacs all depends on what features you prefer.
Task 13.5: Using the emacs Tutorial and Help
System
Unlike vi and, indeed, unlike most of UNIX, emacs includes its own extensive, builtin documentation and a tutorial to help you learn about how to use the program. As
I noted earlier, the entire help system is accessed by pressing C-h. Pressing C-h three times
brings up the general help menu screen. There is also an information browser called info
(accessed by pressing C-h i) and a tutorial system you can start by pressing C-h t.
emacs enthusiasts insist that the editor is modeless, but in fact it does have modes of its own.
You used one just now, the query-replace mode. To obtain help on the current mode that
you’re working in, you can use C-h m.
1. Boldly, I opted to press C-h C-h C-h, and the entire screen is replaced with this:
You have typed C-h, the help character. Type a Help option:
A
B
C
F
I
K
L
command-apropos. Give a substring, and see a list of commands
(functions interactively callable) that contain
that substring. See also the apropos command.
describe-bindings. Display table of all key bindings.
describe-key-briefly. Type a command key sequence;
it prints the function name that sequence runs.
describe-function. Type a function name and get documentation of it.
info. The info documentation reader.
describe-key. Type a command key sequence;
it displays the full documentation.
view-lossage. Shows last 100 characters you typed.
13
Hour 13
298
M
describe-mode. Print documentation of current major mode,
which describes the commands peculiar to it.
N view-emacs-news. Shows emacs news file.
S describe-syntax. Display contents of syntax table, plus explanations
T help-with-tutorial. Select the Emacs learn-by-doing tutorial.
V describe-variable. Type name of a variable;
it displays the variable’s documentation and value.
W where-is. Type command name; it prints which keystrokes
invoke that command.
--**-Emacs: *Help*
(Fundamental)----Top----------------A B C F I K L M N S T V W C-c C-d C-n C-w or Space to scroll: _
What to do now? There are actually 17 different options from this point, as shown
in Table 13.4.
Table 13.4. emacs help system command options.
Command
Meaning
A
List all commands matching the specified word.
List all key mappings.
Describe any key sequence pressed, instead of doing it.
Describe the specified function.
Start up the info browser.
Fully describe the result of a particular key sequence.
Show the last 100 characters you typed.
Describe the current mode you’re in.
List a command syntax table.
Start the emacs tutorial.
Define and describe the specified variable.
Indicate what keystroke invokes a particular function.
Display emacs copyright and distribution information.
Display emacs ordering information.
Display recent emacs changes.
Display emacs warranty.
B
C
F
I
K
L
M
S
T
V
W
C-c
C-d
C-n
C-w
2. I choose K and then press M-< to see what that command really does. The first
thing that happens after typing K is that the table of help information vanishes, to
be replaced by my original text, and then the prompt appears along the bottom:
13
299
of the Danish_people before or during the Revolution, it is truly made,
(Fundamental)----Top----------------Describe key:-
Pressing M-< brings up the desired information:
Preface
----Emacs: dickens.note~
(Fundamental)----Top----------------beginning-of-buffer:
Move point to the beginning of the buffer; leave mark at previous
position.
With arg N, put point N/10 of the way from the true beginning.
Don’t use this in Lisp programs!
(goto-char (point-min)) is faster and does not set the mark.
----Emacs: *Help*
(Fundamental)----All----------------Type C-x 1 to remove help window.
A quick C-x
1
removes the help information when I’m done with it.
There is a considerable amount of help available in the emacs editor. If you’re
interested in learning more about this editor, the online tutorial is a great place to start.
Try C-h t to start it, and go from there.
Task 13.6: Working with Other Files
By this point, it should be no surprise that there are about a million commands
available within the emacs editor, even though it can be a bit tricky to get to them.
There are many file-related commands, too, but I’m going to focus on just a few essential
commands so you can get around in the program. The emacs help system can offer lots more.
(Try using C-h a file to find out what functions are offered in your version of the program.)
13
300
Hour 13
To add the contents of a file to the current edit buffer, use the command C-x i. It will prompt
for a filename. Pressing C-x C-w prompts for a file to write the buffer into, rather than the
default file. To save to the default file, use C-x C-s (that is, if you can; the C-s might again
hang you up, just as it did when you tried to use it for searching). If that doesn’t work, you always
can use the alternative C-x s, which also works. To move to another file, use C-x C-f. (emacs
users never specify more than one filename on the command line. They use C-x C-f to move
between files instead). What’s nice is that when you use the C-x C-f command, you load the
contents of that file into another buffer, so you can zip quickly between files by using the
C-x b command to switch buffers. emacs allows you to edit several files at once using different
areas of the screen; these areas are called buffers.
1. Without leaving emacs, I press C-x C-f to read another file into the buffer. The
system then prompts me as follows:
(Fundamental)----Top----------------Find file: ~/ _
I type buckaroo, and the editor opens up a new buffer, moving me to that file:
I found myself stealing a peek at my own watch and overhead
General Catbird’s
“Banzai.”
13
----Emacs: buckaroo
301
(Fundamental)----All-----------------
2. Now I’ll flip back to the other buffer with C-x b. When I enter that command,
however, it doesn’t automatically move me there. Instead, it offers this prompt:
--**-Emacs: buckaroo
(Fundamental)----All----------------Switch to buffer: (default dickens.note) _
When I type ?, I receive a split screen indicating what the possible answers are:
General Catbird’s
“Banzai.”
Possible completions are:
*Buffer List*
*scratch*
dickens.note
(Fundamental)----All----------------*Help*
buckaroo
13
----Emacs: *Completions*
(Fundamental)----All----------------Switch to buffer: (default dickens.note) _
The default is okay, so I press Return and voila! I’m back in the Dickens file. One
more C-x b; this time the default is buckaroo, so I again press Return to move back.
3. I’m in the buckaroo file, and I want to see what happens if I read dickens.note into
this file. This is done easily. I move the cursor to the end of the file with M->, press
302
Hour 13
C-x i, and answer dickens.note to the prompt Insert
yields the following screen display:
file: ~/ .
Pressing Return
General Catbird’s
“Banzai.”
Preface
care and interest.
(Fundamental)----Top-----------------
5. It’s time to quit and split. To do this, I press C-x
or two. The first one displayed is this:
s
and wait for an emacs prompt
(Fundamental)----Top----------------Save file /users/taylor/buckaroo? (y or n) _
I answer y to save this muddled file. It returns me to the top of the file, and a quick
C-x C-c drops me back to the system prompt.
One of the more useful facets of emacs that you have learned above is the ability to
work with multiple files.
13
303
Summary
You have now learned quite a bit about the emacs editor. Some capabilities exceed those of
the vi editor, and some are considerably more confusing. Which of these editors you choose
is up to you, and your choice should be based on your own preferences for working on files.
You should spend some time working with the editor you prefer, making sure you can create
simple files and modify them without any problems.
Workshop
in the next hour.
Key Terms
Meta key Analogous to a Control key, this is labeled either Meta or Alt on your keyboard.
buffer
An area of the screen used to edit a file in emacs.
named emacs command A command in emacs that requires you to type its name, like
query-replace , rather than a command key or two.
key bindings The emacs term for key mapping.
flow control The protocol used by your computer and terminal to make sure that neither
outpaces the other during data transmission.
XON/XOFF A particular type of flow control. The receiving end can send an XON (delay
transmission) character until it’s ready for more information, when it sends an XOFF
(resume transmission).
Questions
1. How do you get to the emacs help system?
2. Check your keyboard. If you don’t have a Meta or Alt key, what alternative strategy
can you use to enter commands such as M-x?
3. What’s the command sequence for leaving emacs when you’re done?
4. What was the problem I had with the Delete key? How did I solve the problem?
What’s the alternative delete command if Delete isn’t available?
13
304
Hour 13
5. How do you do global search-and-replace in emacs, and what key do you press to
stop the global search-and-replace when you are prompted for confirmation at the
very first match?
6. Use the emacs help system to list the emacs copyright information. What’s your
reaction?
The next hour is an in-depth look at the different shells available in UNIX, how to configure
them, and how to choose which you’d like to use. You also learn about the contents of the
default configuration files for both csh and sh, the two most common shells in UNIX.
13
Hour
305
14
Introduction to
Command Shells
Welcome to your 14th hour of learning UNIX. You should take a moment to
pat yourself on the back. You’ve come a long way, and you’re already quite a
sophisticated user. In the past few hours, I’ve occasionally touched on the
differences between the shells, but I haven’t really stopped to explain what shells
are available, how they differ from one another, and which is the best for your
style of interaction. That’s what this hour is all about.
Shells, you’ll recall, are the command-line interface programs through which
you tell the computer what to do. All UNIX systems include C shell (csh) and
its predecessor, the Bourne shell (sh). Some also include a newer version of the
Bourne shell, called the Korn shell (ksh).
Goals for This Hour
■ What shells are available, and how they differ from one another
■ How to identify which shell you’re running
14
306
Hour 14
■ How to choose a new shell
■ More about the environment of your shell
■ How to explore csh configuration files
A variety of shells are available in UNIX, but two are quite common: the Bourne shell (sh)
and the C shell (csh). You learn about two of the other shells available, the Korn shell (ksh)
and the terminal-based C shell (tcsh). Because the C shell is so popular, most of this book
focuses on it.
Task 14.1: What Shells Are Available?
If I asked a PC expert how many command interpreters are available for DOS, the
immediate answer would be “one, of course.” After a few minutes of reflection,
however, the answer might be expanded to include The Norton Desktop, DesqView,
Windows 95, Windows for Workgroups, and others. This expanded answer reflects the
reality that whenever there are different people using a computer, there will evolve different
styles of interacting with the machine and different products to meet these needs. Similarly,
the Macintosh has several command interpreters. If you decide that you don’t like the
standard interface, perhaps you will find that At Ease, DiskTop, or Square One works better.
From the very beginning, UNIX has been a programmer’s operating system, designed to
allow programmers to extend the system easily and gracefully. It should come as no surprise,
then, that there are quite a few shells available. Not only that, but any program can serve as
a command shell, so you could even start right in emacs if you wanted and then use escapes
to UNIX for actual commands. (Don’t laugh—I’ve heard it’s done sometimes.)
The original shell was written by Ken Thompson, back in the early UNIX’s laboratory days,
as part of his design of the UNIX file system. Somewhere along the way, Steven Bourne, also
at AT&T, got hold of the shell and started expanding it. By the time UNIX began to be widely
distributed, sh was known as the Bourne shell. Characterized by speed and simplicity, it is
the default shell for writing shell scripts, but it is rarely used as a command shell for users
today.
The next shell was designed by the productive Bill Joy, author of vi. Entranced by the design
and features of the C programming language, Joy decided to create a command shell that
shared much of the C language structure and that would make it easier to write sophisticated
shell scripts: the C shell, or csh. He also expanded the shell concept to add command aliases,
command history, and job control. Command aliases allow users to rename and reconfigure
commands easily. Command history ensures that users never have to enter commands a
second time. Job control enables users to run multiple programs at once. The C shell is by
far the most popular shell on all systems I’ve ever used, and it’s the shell that I have been using
myself for about 15 years now, since I first logged in to a BSD UNIX system in 1980.
14
307
A command alias is a shortcut for a command, allowing you to enter a shorter string for the
entire command. The command history is a mechanism by which the shell remembers
commands you have typed and allows you to repeat the command without retyping the whole
command. Job control is a mechanism that allows you to start, stop, and suspend commands.
In the past few years, another AT&T Labs software wizard, David Korn, has begun
distributing another shell on various UNIX platforms. The Korn shell, also known as ksh,
is designed to be a superset of the Bourne shell, sharing its configuration files (.profile) and
command syntax, but including many of the more powerful features of the C shell, too,
including command aliases (albeit in a slightly different format), command history, and job
control. This shell is slowly becoming more popular, but it isn’t yet widely distributed. You
might not have it on your version of UNIX.
Other shells exist in special niches. A modified version of the C shell, a version that
incorporates the slick history-editing features of the Korn shell, has appeared: it is called tcsh.
Maintained by some engineers at Cornell University, it is 95 percent csh and 5 percent new
features. The most important tcsh additions to the C shell are these:
■
■
■
■
■
■
■
■
emacs-style
command-line editing
Visual perusal of the command history list
Interactive command, file, and identifying files with the first few unique characters
Spelling correction of command, file, and user names
Automatic logout after an extended idle period
The capability to monitor logins, users, or terminals
New pre-initialized environment variables $HOST and $HOSTTYPE
Support for a meaningful and helpful system status line
Another shell that you might bump into is called the MH shell, or msh, and it’s designed
around the MH electronic mail program, originally designed at the Rand Corporation. In
essence, the MH shell lets you have instant access to any electronic mail that you might
encounter. For sites that have security considerations, a restricted version of the Bourne shell
is also available, called rsh (ingeniously, it’s called the restricted sh shell). Persistent rumors
of security problems with rsh suggest that you should double-check before you trust dubious
users on your system with rsh as their login shell (The shell you use, by default, when you
log in to the system).
Two other variants of the Bourne shell are worth mentioning: jsh is a version of the Bourne
shell that includes C shell-style job control features, and bash, also humorously called the
“Bourne Again” shell, is a reimplementation of the original shell with many new features and
no licensing restrictions.
14
Hour 14
308
JUST A MINUTE
Licensing restrictions and intellectual property laws occasionally have
stymied the growth of UNIX. Although UNIX is unquestionably popular
with programmers, these same programmers have a burning desire to see
what’s inside, to learn about how UNIX works by examining UNIX itself.
UNIX is owned by AT&T. Few people are able to view the source legally.
Those who do look into UNIX are “tainted”: Anything they write in the
future might be inspired by proprietary code of AT&T. The situation is
fuzzy in many ways, and that’s where the Free Software Foundation
comes in. The brainchild of Richard Stallman, the FSF is slowly rewriting
all the major UNIX utilities and then distributing them with the source, as
part of the ambitious GNU project. GNU emacs is one example, and the
Bourne Again shell is another.
1. In Hour 9 you learned how to use awk to extract the default login shell of each user
on your system and then use sort and uniq to collate the data and present an
attractive output, respectively. Armed with the description of all the different shells,
you now can take another look:
% awk -F: ‘{print $7}’ /etc/passwd | sort | uniq -c
2
3361 /bin/csh
1 /bin/false
85 /bin/ksh
21 /bin/sh
2. You can see that the vast majority of the people on this system use the C shell. To
look at it a different way, compute the number of entries in the password file:
% wc -l /etc/passwd
3849
Now, what percentage of users have chosen each of these shells? This is a job for bc!
% bc
x=3849
scale=4
3361/x*100
87.3200
85/x*100
2.2000
361/x*100
9.3700
^d
It’s a tad difficult to interpret, but this output says that 87.3 percent of the users
have csh as their login shell, another 9.4 percent use the modified tcsh, and only
2.2 percent use ksh. The remaining 1.1 percent use either the Bourne shell or the
14
309
MH shell, or they default to the Bourne shell (the two accounts in the preceding
output, without any shell indicated).
JUST A MINUTE
The scale=4 command tells bc how many digits to display after the
decimal point in numbers. By default, unfortunately, bc displays no digits
after the decimal point.
Quite a variety of shells is available, but the most common one on sites I’m familiar
with is the C shell. Clearly, the system that I used for this particular set of examples
has an overwhelming majority of C shell users: A combined total of 97 percent of the users
are working within either the C shell or its descendent tcsh.
Task 14.2: Identifying Your Shell
There are many different approaches to identifying which shell you’re using. The
easiest, however, is just to swoop into the /etc/passwd file to see what your account
lists. It’s helpful to know some alternatives because the /etc/passwd option isn’t always
available (some systems don’t have an /etc/passwd file in the interest of security).
1. One simple technique to identify your shell is to check your prompt. If your
prompt contains a %, you probably are using the C shell or modified C shell
(tcsh). If your prompt contains $, you could be using the Bourne shell, the Korn
shell, or a variant thereof.
2. A much more reliable way to ascertain which shell you’re using is to ask the
operating system what program you’re currently running. The shell variable $$
identifies the process ID of the shell. You can use $$ as a search pattern for grep on
the output of ps to see what shell you are using. Here’s what happens when I try it:
% ps -ef | grep $$
taylor
26905
0.0
taylor
29751
0.0
0.2
0.1
256
52
144 Ai S
28 Ai S
0:03 -csh (csh)
0:00 grep 26905
You can see that I’m running the C shell. Using ps in this fashion also matches the
grep process (notice that the $$ have expanded to the current shell process identification, 26905). There is a leading dash on the indication of what shell I’m running
because that’s how the system denotes whether it’s my login shell or just a shell that
I’m running.
14
Hour 14
310
3. Another way to find out what shell I’m running is to peek into the /etc/passwd
file, which you can do with some sophistication now that awk is no longer a
mystery:
% awk -F: ‘{ if ($1 == “taylor”) print “your shell is: “$7}’ <
➥/etc/passwd
your shell is: /bin/csh
4. The best way to figure out what shell you’re running, however, is to use chsh. You
learn how to use chsh in the following task.
Once you’ve identified your shell, you can contemplate choosing a different one.
Task 14.3: How To Choose a New Shell
In the past, the only way to switch login shells on many systems was to ask the system
administrator to edit the /etc/passwd file directly. This usually meant waiting until
the system administrator had time. The good news is that there’s now a simple program (on
almost all UNIX systems) to change login shells—it’s chsh, or change shell. It has no starting
flags or options, does not require that any files be specified, and can be used regardless of your
location in the file system. Simply type chsh and press Return.
1. The first step is to identify what shells are available for use. By convention, all shells
have sh somewhere in their names, and they are located in /bin:
% ls -lF /bin/*sh*
-rwsr-xr-x 3 root
-rwxr-xr-x 1 root
-rwxr-xr-x 1 root
-rwxr-xr-x 1 root
49152
102400
139264
28672
Apr 23 1992 /bin/chsh*
Apr 8 1991 /bin/csh*
Jul 26 14:35 /bin/ksh*
Oct 10 1991 /bin/sh*
The chsh command enables you to change your login shell, as you will learn. The
most common shells are csh, ksh, and sh.
On one of the machines I use, some shells are also stored in the /usr/local/bin
directory:
% ls -lF /usr/local/bin/*sh*
lrwxr-xr-x 1 root
8 Jul 26 14:46 /usr/local/bin/ksh ->
➥/bin/ksh*
-rwxr-xr-x 1 root
266240 Jan 19 1993 /usr/local/bin/tcsh*
You can see that there’s an entry in /usr/local/bin for the ksh shell but that it’s
actually just a link pointing to the file in the /bin directory.
14
311
2. You might find quite a few more matches to these simple ls commands. On
another, very different system, I tried the same two commands and found the
following:
% ls -CF /bin/*sh*
/bin/chsh*
/bin/ksh*
/bin/csh*
/bin/sh*
% ls -CF /usr/local/bin/*sh*
/usr/local/bin/bash*
/usr/local/bin/bash112*
/usr/local/bin/ircflush@
/usr/local/bin/mush*
/usr/local/bin/mush.old*
/usr/local/bin/mush725*
/usr/local/bin/ntcsh*
/usr/local/bin/shar*
/usr/local/bin/ship*
/usr/local/bin/showaudio*
/usr/local/bin/showexternal*
/usr/local/bin/shownonascii*
/usr/local/bin/showpartial*
/bin/shelltool@
/bin/shift_lines@
/bin/tcsh*
/bin/ypchsh*
/usr/local/bin/showpicture*
/usr/local/bin/sun-audio-file.csh*
/usr/local/bin/sun-to-mime.csh*
/usr/local/bin/tcsh*
/usr/local/bin/tcsh603*
/usr/local/bin/unshar*
/usr/local/bin/unship*
/usr/local/bin/uupath.sh*
/usr/local/bin/vsh*
/usr/local/bin/zsh*
/usr/local/bin/zsh210*
/usr/local/bin/zsh231*
Two more shells show up here: vsh and zsh. The visual shell, vsh, is an interface
much like the Norton Desktop on DOS. Watch what happens to my screen when
I launch it by typing vsh:
Directory: /u1/taylor
a
b
c
d
e
f
g
h
i
_
User: taylor
Page 2 / 2
.tin/
Global.Software
Interactive.Unix
Mail/
News/
Src/
bin/
history.usenet.Z
testme
14
Hour 14
312
If you have vsh on your system, you might be interested in experimenting with this
very different shell.
The zsh shell is another command shell, one written by Paul Falstad of Princeton
University. The Bourne Again shell, bash, also appears in the listing, as does the
mush program, which is an electronic mail package.
3. Needless to say, many shells are available! To change my login shell to any of these
alternate shells, or even just to verify what shell I’m running, I can use the change
shell command:
% chsh
Old shell: /bin/csh
New shell: _
At this point, the program shows me that I currently have /bin/csh as my login
shell and asks me to specify an alternative shell. I’ll try to confuse it by requesting
that emacs become my login shell:
% chsh
Old shell: /bin/csh
New shell: /usr/local/bin/gnuemacs
/usr/local/bin/gnuemacs is unacceptable as a new shell.
4. The program has some knowledge of valid shell names, and it requires you to
specify one. Unfortunately, it doesn’t divulge that information, so typing ? to find
what’s available results in the program complaining that ? is unacceptable as a new
shell.
You can, however, peek into the file that chsh uses to confirm which programs are
valid shells. It’s called /etc/shells and looks like this:
% cat /etc/shells
/bin/ksh
/bin/sh
/bin/csh
/usr/local/bin/ksh
/usr/local/bin/tcsh
I’ll change my shell from /bin/csh to /usr/local/bin/tcsh:
% chsh
Changing login shell for taylor
Old shell: /bin/csh
New shell: /usr/local/bin/tcsh
Notice that, in typical UNIX style, there is no actual confirmation that anything
was done. I conclude that, because I did not get any error messages, the program
worked. Fortunately, I easily can check by either using chsh again or redoing the
awk program with a C shell history command:
14
% !awk
awk -F: ‘{ if ($1 == “taylor”) print “your shell is: “$7}’ < /etc/passwd
your shell is: /usr/local/bin/tcsh
313
In the next hour, you learn more about the powerful C shell command-history
mechanism.
JUST A MINUTE
Because of the overwhelming popularity of the C shell, the next few hours
focus on the C shell. To get the most out of those hours, I strongly recommend that you use the C shell.
5. A quick reinvocation of the chsh command changes my shell back to /bin/csh:
% chsh
Changing login shell for taylor
Old shell: /usr/local/bin/tcsh
New shell: /bin/csh
If you can’t change your login shell, perhaps because of not having chsh,
you always can enter the C shell after you log in by typing csh.
JUST A MINUTE
It’s easy to change your login shell. You can try different ones until you find the one
that best suits your style of interaction. For the most part, though, shells all have the
same basic syntax and use the same commands: ls -l does the same thing in any shell. The
differences, then, really come into play when you use the more sophisticated capabilities,
including programming the shell (with shell scripts), customizing its features through
command aliases, and saving on keystrokes using a history mechanism. That’s where the C
shell has an edge and why it’s so popular. It’s easy, straightforward, and has powerful aliasing,
history, and job-control capabilities, as you learn in the next hour.
Task 14.4: Learning the Shell Environment
Earlier in this book, you used the env or printenv command to find out the various
characteristics of your working environment. Now it’s time to use this command
again to look more closely at the C shell environment and define each of the variables therein.
1. To start out, I enter env to list the various aspects of my working environment. Do
the same on your system, and, although your environment will not be identical to
mine, there should be considerable similarity between the two.
% env | cat -n
1 HOME=/users/taylor
2 SHELL=/bin/csh
3 TERM=vt100
14
Hour 14
314
4 USER=taylor
5 PATH=.:/users/taylor/bin:/bin:/usr/bin:/usr/ucb:/usr/local:/etc:
➥/usr/etc:/usr/local/bin:/usr/unsup/bin:
6 MAIL=/usr/spool/mail/taylor
7 LOGNAME=taylor
8 EDITOR=/ucb/bin/vi
9 NAME=Dave Taylor
10 EXINIT=:set ignorecase
11 RNINIT=-hmessage -hreference -hdate-r -hsender -hsummary -hreply
➥-hdistr -hlines -hline -hfollow -hnews -hkey -hresent -hreturn -hto
➥-hx-original -hx-sun -hx-note -horiginator -hnntp
12 SUBJLINE=%t -- %s
13 ORGANIZATION=Educational Computing group, School of Education
This probably initially seems pretty overwhelming. What are all these things, and
why on earth should they matter? They matter because it’s important for you to
learn exactly how your own environment is set up so that you can change things if
you desire. As you soon will be able to recognize, I have modified much of my
system’s environment so that the C shell does what I want it to do, rather than
what its default would tell it to do.
2. When I log in to the system, the system defines some environment variables,
indicating where my home directory is located, what shell I’m running, and so on.
These variables are listed in Table 14.1.
Table 14.1. Default variables set by UNIX upon login.
Variable
Description
HOME
This is my home directory, obtained from the fourth field of the
password file. Try the command
grep $USER /etc/passwd | awk -F: ‘{print $6}’
SHELL
TERM
14
to see what your home directory is set to, or just use env HOME or echo
$HOME. This is not only the directory that I start in, but it’s also the
directory that cd moves me back to when I don’t specify a different
directory. My HOME variable is /users/taylor.
When UNIX programs, such as vi, process the ! command to execute
UNIX commands, they check this variable to see which shell I’m
using. If I were to type :! followed by Return in vi, the program
would create a new C shell for me. If I had SHELL=/bin/sh, vi would
start up a Bourne shell. My SHELL variable is set to /bin/csh.
By default, your terminal is defined by the value of this environment
variable, which starts out as unknown. (Recall that when you first were
learning about vi, the program would complain unknown: terminal
not known.) Many sites know what kind of terminals are using which
lines, however, so this variable is often set to the correct value before
you even see it. If it isn’t set, you can define it to the appropriate value
within your .login file. (You will learn to do this later in the hour.)
Variable
USER
PATH
315
Description
My TERM is set to vt100, for a Digital Equipment Corporation Visual
Terminal model 100, which is probably the most commonly emulated
terminal in communications packages.
Programs can quickly look up your user ID and match it with an
account name. However, predefining your account name as an
environment setting saves time. That’s exactly what USER, and its
companion LOGNAME, are—timesavers. My USER is set to taylor.
A few hours ago, you learned that the UNIX shell finds a command by
searching from directory to directory until it finds a match. The
environment variable that defines which directories to search and the
order in which to search them is the PATH variable. Rather than keep
the default settings, I’ve added a number of directories to my search
path, which is now as follows:
.:/users/taylor/bin:/bin:/usr/bin:/usr/ucb:/usr/local:/etc
:/usr/etc:/usr/local/bin:/usr/unsup/bin:
I have told the shell always to look first for commands in the current
directory (.), then in my bin directory (/users/taylor/bin), and then
in the standard system directories (/bin, /usr/bin, /usr/ucb, /usr/
local). If the commands are not found in any of those areas, the shell
should try looking in some unusual directories (/etc, /usr/etc, /usr/
local/bin). If the shell still has not found my command, it should
check in a weird directory specific to my site: /usr/unsup/bin for
unsupported software, /usr/unsup/elm for programs related to the
Elm Mail System, and /usr/local/wwb for the AT&T Writers Workbench programs.
CAUTION
I admit it; using . as the first entry in the PATH variable is a security
hazard. Why? Imagine this: A devious chap has written a program that
will do bad things to my directory when I invoke that bad program. But
how will he make me invoke it? The easiest way is to give the bad
program the same name as a standard UNIX utility, such as ls, and leave
it in a commonly accessed directory, such as /tmp. So what happens?
Imagine that the . (current directory) is the first entry in my PATH, and I
change directories to /tmp to check something. While I’m in /tmp, I enter
ls without thinking, and voilà! I’ve run the bad program without knowing
it. Having the . at the end of the search path would avoid all this because
then the default ls command is the correct version. I have it because I
often do want to override the standard commands with new ones that I’m
working on (an admittedly foolish practice).
continues
14
Hour 14
316
Variable
MAIL
LOGNAME
JUST A MINUTE
NAME
Description
One of the most exciting and enjoyable aspects of UNIX is its powerful and incredibly well-connected electronic mail capability. A variety
of programs can be used to check for new mail, to read mail, and to
send mail messages. Most of these programs need to know where my
default incoming mailbox is located, which is what the MAIL environment variable defines. My MAIL is set to /usr/spool/mail/taylor.
LOGNAME is a synonym for USER. My LOGNAME is set to taylor.
Having both LOGNAME and USER defined in my environment demonstrates
how far UNIX has progressed since the competition and jostling between
the Berkeley and AT&T versions (BSD and SVR3, respectively) of UNIX.
Back when I started working with UNIX, if I was on a BSD system, the
account name would be defined as LOGNAME, and if I used an SVR3
system, the account name would be defined as USER. Programs had to
check for both, which was frustrating. Over time, each system has begun
to use both terms (instead of using the solution that you and I might think is
most obvious, which is to agree on a single word).
In addition to wanting to know the name of the current account, some
programs, such as many electronic mail and printing programs, need
to ascertain my full, human name. The NAME variable contains this
information for the environment. It’s obtained from the /etc/passwd
file. You can check yours with the command grep $USER /etc/passwd
| awk -F: ‘{print $5}’. You can change your NAME variable, if
desired, using the chfn, or change-full-name, command. My NAME is set
to Dave Taylor.
3. A glance back at the output of the env command reveals that there are more
variables in my environment than are listed in Table 14.1. That’s because you can
define anything you want in the environment. Certain programs can read many
environment variables that customize their behavior.
Many UNIX programs allow you to enter text directly, and then they spin off into
an editor, if needed. Others start your favorite editor for entering information.
Both types of programs use the EDITOR environment variable to identify which
editor to use. I have mine set to /usr/ucb/vi.
14
317
You learned earlier that vi can have default information stored in the .exrc file, but
the program also can read configuration information from the environment
variable EXINIT. To make all my pattern searches case insensitive (that is, searching
for precision will match Precision), I set the appropriate vi variable in the EXINIT.
Mine is set to :set ignorecase. If you want line numbers to show up always, you
could easily have your EXINIT set to :set number.
Another program that I use frequently is rn, or read Netnews. If electronic mail is
the electronic equivalent of letters and magazines that you receive through the
postal service, Netnews is the electronic equivalent of a super bulletin board. The
difference is that there are thousands of different boards, and any time a note is
tacked onto any board, copies of the note shoot to other UNIX systems throughout
the world. For now, you can see that I have three environment variables all defined
for the rn program: RNINIT, my personal rn configuration options; SUBJLINE,
indicating the format for displaying summary subject lines of new messages; and
ORGANIZATION , indicating exactly what organization I’m associated with on this
system. They are set as shown earlier.
There are many possible environment variables that you can define for yourself. Most
large UNIX programs have environment variables of their own, allowing you to tailor
the program’s behavior to your needs and preferences. UNIX itself has quite a few
environment variables, too. Until you’re an expert, however, I recommend that you stick
with viewing these variables and ensuring that they have reasonable values, rather than
changing them. Particularly focus on the set of variables defined in Table 14.1. If they’re
wrong, it could be trouble; whereas, if other environment variables are wrong, it’s probably
not going to be too much trouble.
Task 14.5: Exploring csh Configuration Files
The C shell uses two files to configure itself, and, although neither of them need to
be present, both probably can be found in your home directory: .login and .cshrc.
The difference between them is subtle but very important. The .login file is read only once,
when you log in, and the .cshrc file is read every time a C shell is started. As a result, if you’re
working in vi and you enter :!ls, vi carries out the command by starting up a new shell and
then feeding the command to that shell. Therefore, new csh shells started from within
programs such as vi won’t see key shell configurations that are started in .login.
This split between two configuration files isn’t too bad, actually, because many modifications
to the environment are automatically included in all subshells (a shell other than the login shell)
invoked. To be specific, all environment variables are pervasive, but any C shell command
aliases are lost and, therefore, must be defined in the .cshrc file to be available upon all
occurrences of csh. You learn more about command aliases in the C shell in the next hour.
14
Hour 14
318
1. To begin, I use cat to list the contents of my .login file. Remember that any line
beginning with a # is a comment and is ignored.
% cat .login
#
# @(#) $Revision: 62.2 $
setenv TERM vt100
stty erase “^H” kill “Û” intr “^C” eof “^D”
stty crtbs crterase
# special DYNIX stuff for bs processing
# shell vars
set noclobber history=100 savehist=50 filec
# set up some global environment variables...
setenv EXINIT “:set ignorecase”
# Some RN related variables...
setenv RNINIT
“-hmessage -hreference -hdate-r -hsender -hsummary
-hreply -hdistr -hlines -hline -hfollow -hnews -hkey -hresent -hreturn
-hto -hx-original -hx-sun -hx-note -horiginator -hnntp”
setenv SUBJLINE
“%t — %s”
setenv ORGANIZATION
“Educational Computing group, School of Education
“
setenv NAME “Dave Taylor”
newmail
mesg y
This is pretty straightforward, once you remove all the comments. Three different
kinds of environmental configuration commands are shown: setenv, stty, and set.
The setenv command defines environment variables; indeed, you can see that
many of the variables shown in the previous unit are defined in my .login file.
I can use stty commands to set specific configuration options related to my
terminal (stty stands for “set tty driver options”). I use this to ensure that ^h is
erase (backspace), û is a convenient shortcut allowing me to kill an entire line, and
^c sends an interrupt to a running program. I indicate the end of a file (EOF) with
^d. The second line of the preceding output example indicates that my CRT is
capable of backspacing and erasing characters on the display.
14
JUST A MINUTE
319
Here’s more arcane UNIX nomenclature: CRT (as used in stty crtbs or
cathode-ray tube) is the technology used in the screen of a standard
terminal. Terminal is not accurate anymore, however, because the
command stty crtbs also works on my LCD (liquid-crystal diode, if you
must know) laptop.
Finally, the set commands are configuration options for the C shell. I have told the
C shell to warn me before it overwrites existing files with file redirection
(noclobber), to remember the last 100 commands (history=100), and to remember
50 of those even if I log out and log back in (savehist=50). I also want the C shell
to try, if possible, to complete filenames for me, hence the filec addition. Notice
that there are two different types of settings: on/off options (such as noclobber and
filec) and options to which I must assign a specific numeric value (such as history
and savehist).
The two commands at the very end of the .login file are invoked as though I’d
entered them on the command line. The newmail variable watches for new electronic mail (in the mailbox defined by the environment variable MAIL, in fact) and
tells me when it arrives. The mesg y variable makes sure that I have my terminal
configured so that other folks can beep me or say hello using write or talk, two
communication tools discussed in Hour 20, “Communicating with Others.”
2. How about the other file—the one that’s read by the C shell each time a shell is
started?
% cat .cshrc
#
set path=(. ~/bin /bin /usr/bin /usr/ucb /usr/local /etc /usr/etc
/usr/local/bin /usr/unsup/bin /
# Define a bunch of C shell aliases
alias
alias
alias
alias
alias
tn3270
alias
alias
alias
alias
alias
diff
env
‘printenv’
from
‘frm -n’
info
ssinfo
library
‘echo “ “; echo “ “ ; echo “remember: ^J is ENTER”;
lib
ll
‘ls -l’
ls
‘/bin/ls -F’
mail
Mail
mailq
14
Hour 14
320
alias
alias
alias
rd
rn
ssinfo
alias
alias
cd
setprompt
set noclobber history=100 system=mentor filec
umask 007
setprompt
endif
Again, any line that begins with a # is considered a comment. There are, therefore,
two primary types of commands in this script: the C shell environment modification (set) and the command alias (alias). The first defines the PATH I want to use,
although in a format slightly different from the colon-separated list shown by env.
The csh command always ensures that the environment variable and shell variable
match, and so, although I opt to change the path here as a set, I could just as easily
use setenv PATH.
You learn all about aliases in the next hour, but for now you should know that the
format is alias word command (or commands) to execute. When I enter ls, for
example, you can see that the shell has that aliased to /bin/ls -F, which saves me
from having to type the -F flag each time.
The C shell also has conditional statements and a variety of other commands to
indicate what commands to run. Here I’m using the if (expression) then to
define a set of commands that should be used only when the shell is interactive
(that is, I’m going to be able to enter commands). An example of a noninteractive
shell is the shell vi uses to create a listing when I enter !!ls within the editor. The
$?prompt variable is true if there is a prompt defined for the shell (that is, if it’s
interactive). If not, the variable is false, and the shell zips to the endif before
resuming execution of the commands.
If it is an interactive shell, however, I create a few further aliases and again define
some C shell configuration options, to ensure that the options are always set in
subshells. The umask value is set, and I then invoke setprompt, which is a command alias that runs the command set prompt=”$system ($cwd:t) \! : “.
If you’re thinking that there is a variety of ways to configure the shell, you are correct.
You can have an incredibly diverse set of commands in both your .login and .cshrc
files, enabling you to customize many aspects of the C shell and the UNIX environment. If
you use either the Bourne shell or the Korn shell, the configuration information is kept in
a similar file called .profile.
14
321
Summary
Armed with the information learned in this hour about shells and shell environments, explore
your own environment; examine your .login and .cshrc files, too.
Workshop
in the next hour.
Key Terms
command alias A shorthand command mapping, with which you can define new
command names that are aliases of other commands or sequences of commands. This is
helpful for renaming commands so that you can remember them, or for having certain flags
added by default.
command history A mechanism the shell uses to remember what commands you have
entered already, and to allow you to repeat them without having to type the entire command
again.
job control A mechanism for managing the various programs that are running. Job control
enables you to push programs into the background and pull them back into the foreground
as desired.
login shell The shell you use, by default, when you log in to the system.
subshell A shell other than the login shell.
Questions
1. Draw lines to connect the original shells with their newer variants:
sh
ksh
tcsh
csh
2.
3.
4.
5.
6.
What does chsh do? What about chfn?
What shell are you running? What shells are your friends on the system running?
What’s the difference between the .login and the .cshrc files?
What’s the sh equivalent of the csh .login file?
What aliases do you think could prove helpful for your daily UNIX interaction?
14
322
Hour 14
I hope this hour has whetted your appetite for learning more about the C shell! In the next
hour, you learn how to really customize the shell and make your interaction with UNIX quite
a bit easier. Topics include how to create command aliases, how to use the history mechanism,
and how to create simple shell scripts when aliases just don’t suffice.
14
2
3
15
Hour
15
Getting the Most Out of
the C Shell
The previous hour gave you an overview of the different shells available in
UNIX. There are quite a few, but the C shell—originally from Berkeley,
California—is the most popular shell at most sites. In this hour, you learn all
about the C shell and how to use it to your best advantage. You also learn some
valuable tips about working with the Korn shell, a popular alternative to C shell.
The goal is for you to be able to customize your UNIX environment to fit your
working style.
Goals for This Hour
■
■
■
■
How to turn on the C shell and Korn shell history mechanism
How to use csh history and ksh history to cut down on typing
About command aliases in the C and Korn shells
Some power aliases for csh
324
Hour 15
■ How to set custom prompts
■ How to create simple shell scripts
This hour focuses on two key facets of the C and Korn shells: the history mechanism and the
command alias capability. I guarantee that within a few minutes of learning about these two
functions, it will be clear that you couldn’t have survived as happily in UNIX without them.
There are three ways to ensure that you don’t enter commands more than once: csh history
enables you to repeat previous commands without re-entering them, an alias enables you to
name one command as another, and shell scripts enable you to toss a bunch of commands
into a file to be used as a single command. You learn the basics about building shell scripts
in this hour.
One of the fun parts of UNIX is that you can customize the prompt that greets you each time
you use the system. There’s no need to be trapped with a boring % prompt anymore!
Task 15.1: The C Shell and Korn Shell History
Mechanisms
If you went through school in the United States, you doubtless have heard the
aphorism “Those who do not study history are doomed to repeat it.” UNIX stands
this concept on its head. For UNIX, the aphorism is best stated, “Those who are aware of their
history can easily repeat it.”
Both the C shell and the Korn shell build a table of commands as you enter them and assign
them a command number. Each time you log in, the first command you enter is command
1, and the command number is incremented for each subsequent command that you enter.
You can review or repeat any previous command easily with just a few keystrokes.
To review your history in the C shell, enter history at the csh prompt. Odds are that nothing
will happen, though, because by default csh remembers only the very last command. To have
it begin building a list of commands, you must turn on this feature through an environment
setting set history=n, where n is the number of commands you’d like it to recall.
By contrast, ksh has a default history list size of 128 commands, plenty for anyone. To review
your history in ksh, you also can use the history command. Actually, though, it’s an alias,
and the real command is the more cryptic fc -l. In Korn shell, you don’t need to make any
changes; the history mechanism is ready to use immediately.
1. Log in to your system so that you have a C shell prompt. If you’re currently in the
Bourne shell, this would be a great time to use chsh to change shells.
15
5
2
3
% history
%
The shell indicates that it has no history. Sir Winston Churchill doubtless would
shake his head and mutter under his breath, “To have become such a sophisticated
operating system yet never to have studied history!”
2. I need to turn on the shell history mechanism, so I will enter the following command:
% set history=100
Still there is no feedback, but I can check the status of all the shell parameters by
entering set at the csh prompt:
% set
argv
()
cwd
/users/taylor
filec
history 100
home
/users/taylor
host
limbo
noclobber
path
(. /users/taylor/bin /bin /usr/bin /usr/ucb /usr/local /etc
/usr/etc /usr/local/bin /usr/unsup/bin
savehist
50
shell
/bin/csh
status 0
system limbo
term
unknown
user
taylor
3. To take a break, I use w to see who is logged on and what they’re doing, ls to check
my files again, and date to see if my watch is working:
% w | head
11:41am up 17:59, 103 users,
User
tty
login@ idle
root
console
6:02pm 12:13
taylor
ttyAf
11:40am
bev
ttyAg
9:25am
14
rekunkel ttyAh
11:37am
gabh
ttyAi
10:41am
6
af5
ttyAj
8:27am
21
techman ttyAk
9:47am
tuccie
ttyAl
11:37am
Broken pipe
% ls
Archives/
OWL/
InfoWorld/
awkscript
Mail/
bin/
News/
buckaroo
% date
Tue Dec 7 11:41:29 EST 1993
load average:
JCPU
PCPU
1
1
5
2
1:09
3
4
3
46
16
33
1
25
7
1
1
0.54, 0.53, 0.49
what
-csh
w
-csh
rlogin ccn
talk dorits
-ksh
gopher
mail
buckaroo.confused
dickens.note
keylime.pie
owl.c
sample
sample2
src/
temp/
15
Hour 15
326
JUST A MINUTE
Notice that at the end of the w command output, the system noted Broken
pipe. This is nothing to be anxious about; it’s just an indication that there
was lots more in the pipeline when the program quit. You can see that
head read only the first 10 lines. The first line of the w output shows that
there are 103 users on the system, which means that head ignored 94
lines of output. Unlike real plumbing, fortunately, this broken pipe doesn’t
allow the spare data to spill onto the basement floor!
4. Now, when I enter history, the shell remembers the previous commands, presenting them all in a neat, numbered list:
% history
1 set history=100
2 w | head
3 ls
4 date
5 history
5. To turn this on permanently, add the set history command to your .cshrc. If
you want the shell to remember commands even if you log out and log back in,
also specify the setting savehist. I choose to do this by entering vi +$ .cshrc and
adding the following line:
set noclobber system=limbo filec
umask 007
setprompt
endif
set history=100 savehist=50 _
~
~
If you glance back at the output of the set command, you can see that I already
have both of these parameters set: 100 commands will remain in the history list
while I’m working, and 50 commands will be retained for the next time I log in.
What’s particularly helpful is that any time I specify a number n for either history
list, the shell actually saves the most recent n commands, so I have the most recent
100 commands for review while I’m using the system, and the 50 most recent
commands remembered when I log in later.
Make this change to your .cshrc file, log out, and log in again to ensure that your
history mechanism is set up correctly.
15
Like much of UNIX, turning on the history mechanism of the C shell is quite easy
once you learn the trick. In this case, simply remember that you need to specify a set
history value to have the shell begin remembering what’s going on with your interaction.
In Korn shell, you don’t need to make any changes; it’s ready to use immediately.
Task 15.2: Using History to Cut Down on Typing
There are three main mechanisms for working with the history list. You can specify
a previous command by its command number, by the first word of the command,
or, if you’re working with the most recently executed command, by a special notation that
easily fixes any mistakes you might have made as you typed it.
Every history command begins with an exclamation point. If the 33rd command you entered
was the who command, for example, you can execute it by referring to its command number:
enter !33 at the command line. You can execute it also by entering one or more characters
of the command: !w, !wh, or !who: You must enter enough characters to uniquely identify it
in the history list.
To edit a previous command, type a caret, the pattern you want to change, another caret, and
the correct pattern. If you just entered awk -F, ‘{print $2}’ and realize that you meant to
type a colon, not a comma, as the field delimiter, ^,^: will do the trick.
A very useful shorthand is !!, which repeats the most recently executed command. Two other
history references are valuable to know: !$ expands to the last word of the previous line (which
makes sense because $ always refers to the end of something, whether it be a line, the file, or,
in this case, a command), and !* expands to all the words in the previous command except
the very first. So, for example, if I entered the command ls /usr /etc /dev and then
immediately entered the command echo !*, the second command would be expanded
automatically to echo /usr /etc /dev.
Korn shell offers all of this and more. You can repeat commands by number by specifying
rn, where n is the command number (for example, r33). You can also repeat by name with
rname, as in rwho to repeat the most recent who command. Much more useful is the ksh
capability to edit directly a command with the familiar vi or EMACS command keys,
without leaving the command line. Without any arguments, r will repeat the previous
command.
1. First, I need to spend a few minutes building up a history list by running various
commands:
% w | head
11:58am up 18:14, 81 users, load average: 0.54, 0.44, 0.38
User
tty
login@ idle
JCPU
PCPU what
7
2
3
15
Hour 15
328
root
console
6:02pm 12:30
1
1 -csh
hopkins ttyAe
11:49am
4
4 telnet whip.isca.uiowa.edu
taylor
ttyAf
11:40am
8
2 wbev
ttyAg
9:25am
31
1:09
3 -csh
af5
ttyAj
8:27am
37
33
1 -ksh
techman ttyAk
9:47am
4
1:11
4 elm
tuccie
ttyAl
11:37am
2
1 mail
trice
ttyAm
8:16am 1:21
5
2 -csh
Broken pipe
% date
Tue Dec 7 11:58:19 EST 1993
% ls
Archives/
OWL/
buckaroo.confused sample
InfoWorld/
awkscript
dickens.note
sample2
Mail/
bin/
keylime.pie
src/
News/
buckaroo
owl.c
temp/
% cat buckaroo
I found myself stealing a peek at my own watch and overheard General
Catbird’s
"Who?"
“Banzai.”
%
2. Now I will check my history list to see what commands were squirreled away for
later:
% history
51 set history=100
52 history
53 w | head
54 date
55 ls
56 cat buckaroo
57 history
JUST A MINUTE
I already have my history mechanism turned on, so my commands begin
numbering with 51 rather than with 1. Your system might be different.
Regardless of what the command numbers are, they’ll work!
3. To repeat the date command, I can specify its command number:
% !54
date
Tue Dec
7 12:04:08 EST 1993
Notice that the shell shows the command I’ve entered as command number 54.
The ksh equivalent here would be r 54.
15
9
2
3
4. A second way to accomplish this repeat, a way that is much easier, is to specify the
first letter of the command:
% !w
w | head
12:05pm up 18:23, 87 users, load average: 0.40, 0.39, 0.33
User
tty
login@ idle
JCPU
PCPU what
root
console
6:02pm 12:37
1
1 -csh
lloyds
ttyAb
12:05pm
1
1 mail windberg
lusk
ttyAc
12:03pm
3
2 gopher
hopkins ttyAe
11:49am
8
8 telnet whip.isca.uiowa.edu
taylor
ttyAf
11:40am
1
14
3 w
bev
ttyAg
9:25am
38
1:09
3 -csh
libphar ttyAh
12:03pm
3
3 elm
dgrove
ttyAi
12:02pm
5
2 more inbox/16
Broken pipe
5. Now glance at the history list:
% history
51 set history=100
52 history
53 w | head
54 date
55 ls
56 cat buckaroo
57 history
58 date
59 w | head
60 history
Commands expanded by the history mechanism are stored as the expanded
command, not as the history command that actually was entered. Thus, this is an
exception to the earlier rule that the history mechanism always shows what was
previously entered. It’s an eminently helpful exception!
History commands are quite helpful for people working on a software program.
The most common cycle for programmers to repeat is edit-compile-run, over and
over again. The commands UNIX programmers use most often probably will look
something like vi test.c, cc -o test test.c, and test, to edit, compile, and run
the program, respectively. Using the C shell history mechanism, a programmer
easily can enter !v to edit the file, !c to compile it, then !t to test it. As your
commands become longer and more complex, this function proves more and more
helpful.
6. It’s time to experiment a bit with file wildcards.
% ls
Archives
InfoWorld
Mail
News
OWL
awkscript
bin
buckaroo
buckaroo.confused
cshrc
dickens.note
keylime.pie
owl.c
sample
sample2x
src
temp
15
Hour 15
330
Oops! I meant to specify the -F flag to ls. I can use !! to repeat the command,
then I can add the flag:
% !! -F
ls -F
Archives/
InfoWorld/
Mail/
News/
OWL/
JUST A MINUTE
JUST A MINUTE
awkscript
bin/
buckaroo
buckaroo.confused
cshrc
dickens.note
keylime.pie
owl.c
sample
sample2
src/
temp/
The general idea of all these history mechanisms is that you specify a
pattern that is replaced by the appropriate command in the history list. So,
you could enter echo !! to have the system echo the last command, and it
would end up echoing twice. Try it.
Korn shell users will find that echo !! produces !! and that the ksh repeatlast-command of r also will fail. If your last command was echo r, the
result will be r. Further, there is no analogous shorthand to the convenient
!! -F in csh. On the other hand, if FCEDIT is set to vi or EMACS, you can
pop into the editor to change the command by typing fc.
I want to figure out a pattern or two that will let me specify both buckaroo files, the
dickens file, and sample2, but not sample. This is a fine example of where the echo
command can be helpful:
% echo b* d* s*
bin buckaroo buckaroo.confused dickens.note sample sample2 src
That’s not quite it. I’ll try again:
% echo bu* d* sa*
buckaroo buckaroo.confused dickens.note sample sample2
That’s closer. Now I just need to remove the sample file:
% echo bu* d* sa*2
buckaroo buckaroo.confused dickens.note sample2
That’s it. Now I want to compute the number of lines in each of these files. If I use
the csh history mechanism, I can avoid having to enter the filenames again:
% wc -l !*
wc -l bu* d* sa*2
36 buckaroo
11 buckaroo.confused
28 dickens.note
15
1
3
4 sample2
79 total
Notice that the !* expanded to the entire previous command except the very
first word.
7. What happens if I use !$ instead?
% wc -l !$
wc -l sa*2
4 sample2
8. In the middle of doing all this, I became curious about how many people on my
system have first names that are four letters long. Is this impossible to compute?
Not with UNIX!
The first step is to extract the full names from the /etc/passwd file:
% awk -F: ‘{ print $5 }’
The system does not respond. I forgot to specify the filename!
% !! < /etc/passwd
awk -F: ‘{print $5}’ < /etc/passwd
limbo root,,,,
USENET News,,,,
INGRES Manager,,,,
(1000 user system) DO NOT,,,,
Vanilla Account,,,,
The Ferryman,,,,
I can use ^c to stop this output, because I’ve seen enough to know that it’s what I
want. Next, I use awk again to pull just the first names out of this list:
% !! | awk ‘{print $1}’
awk -F: ‘{print $5}’ < /etc/passwd | awk ‘{print $1}’
root
USENET
INGRES
(1000
Vanilla
The
Account
^c
%
It looks okay. Now the final step: I need to revise this awk script to look at the
length of each name, and output the name only if it’s four letters long:
% !-2 | awk ‘{ if (lng($1) == 4) print $0 }’
awk -F: ‘{print $5}’ < /etc/passwd | awk ‘{ if (lng($1) == 4) print $0 }’
I got no output at all! The reason is that I mistyped length and lng. Fortunately, to
fix this is simplicity itself with C shell history commands. Remember, the format is
ôld^new:
% ^lng^length
awk -F: ‘{print $5}’ < /etc/passwd | awk ‘{ if (length($1) == 4) print $0
15
Hour 15
332
}’
,,,,
Paul
Pete
John
Dana
Dick
Mike
Bill
Dale
Bill
Gary
Doug
Ruth
Dave
^c
%
Town,,,,
Cheese,,,,
Smith,,,,
Tott,,,,
Ply,,,,
Moliak,,,,
Born,,,,
Tott,,,,
Rison,,,,
Flint,,,,
Sherwood,,,,
Raffy,,,,
Sean,,,,
That’s very close. I just need to pipe the output of this command to wc:
% !! | wc -l
awk -F: ‘{print $5}’ < /etc/passwd | awk ‘{ if (length($1) == 4) print
$0 }’ | wc -l
723
9. If you are using Korn shell, here’s where it shines! Make sure that the environment
variable EDITOR is set to your preferred editor:
$ echo $EDITOR
vi
$
Now, any time you’re entering a command, you can press the Escape key and be in
ksh history-edit command mode. The usual vi commands work, including h and l
to move left and right; i and Escape to enter and leave insert mode; w, W, b, and B to
zip about by words; and 0 and $ to move to the beginning or end of the line.
Much more useful are k and j, which replace the current command with the
previous or next, enabling you to zip through the history list.
If I’d just entered who and then ls, to append | wc -l to the who command, I could
press the Escape key:
$_
Now each time I type k, I will see the previous command. Typing k once reveals
this:
$ls
Typing k as second time reveals this:
$who
That’s the right command, so $ moves the cursor to the end of the line:
$who
15
Typing a appends, at which point I can add |
wc -l
like this:
$who | wc -l
Pressing Return results in ksh actually executing the command:
$ who | wc -l
130
$_
The history mechanisms of the shells are wonderful timesavers when you’re working
with files. I find myself using the csh !! and !word mechanisms daily either to build
up complex commands (such as the previous example, in which I built up a very complex
command, step by step) or to repeat the most recently used edit commands. Table 15.1
summarizes the different csh history mechanisms available. I encourage you to learn and use
them. They soon will become second nature and will save you lots of typing.
Table 15.1. C shell history commands.
Command
Function
!!
Repeat the previous command.
Repeat the last word of the previous command.
Repeat all but the first word of the previous command.
Replace a with b in the previous command.
Repeat command n from the history list.
!$
!*
â^b
!n
3
Task 15.3: Command Aliases
If you think the history mechanism has the potential to save you typing, you just
haven’t learned about the command-alias mechanism in the Korn and C shells.
Using aliases, you easily can define new commands that do whatever you’d like, or even
redefine existing commands to work differently, have different default flags, or more!
The general format for using the alias mechanism in csh is alias word command-sequence ,
and in ksh it is alias word=commands. If you enter alias without any specified words, the
output shows a list of aliases you have defined. If you enter alias word in csh, the output lists
the current alias, if there is one, for the specified word.
1. One of the most helpful aliases you can create specifies certain flags to ls so that
each time you enter ls, the output will look as though you used the flags with the
command. I like to have the -FC flags set.
15
Hour 15
334
% ls
Archives
InfoWorld
Mail
News
OWL
awkscript
bin
buckaroo
buckaroo.confused
cshrc
dickens.note
keylime.pie
owl.c
sample
sample2
src
temp
Now I’ll try to create a C shell alias and try it again:
% alias ls ‘ls -CF’
% ls
Archives/
awkscript
InfoWorld/
bin/
Mail/
buckaroo
News/
buckaroo.confused
OWL/
cshrc
dickens.note
keylime.pie
owl.c
sample
sample2
src/
temp/
This is very helpful!
The ksh equivalent would be alias ls = ‘ls -CF’.
2. If you’re coming from the DOS world, you might have found some of the UNIX
file commands confusing. In DOS, for example, you use DIR to list directories, REN
to rename files, COPY to copy them, and so on. With aliases, you can recreate all
those commands, mapping them to specific UNIX equivalents:
% alias DIR ‘ls -lF’
% alias REN ‘mv’
% alias COPY ‘cp -i’
% alias DEL ‘rm -i’
% DIR
total 33
drwx------ 2 taylor
drwx------ 3 taylor
drwx------ 2 taylor
drwx------ 2 taylor
drwx------ 4 taylor
-rw-rw---- 1 taylor
drwx------ 2 taylor
-rw-rw---- 1 taylor
-rw-rw---- 1 taylor
-rw------- 1 taylor
-rw-rw---- 1 taylor
-rw-rw---- 1 taylor
-rw-rw---- 1 taylor
-rw-rw---- 1 taylor
-rw-rw---- 1 taylor
drwx------ 2 taylor
drwxrwx--- 2 taylor
% COPY sample newsample
%
512
512
1024
512
532
126
512
1393
458
1339
1123
12556
8729
199
207
512
512
Nov
Dec
Dec
Oct
Dec
Dec
Oct
Dec
Dec
Dec
Dec
Nov
Dec
Dec
Dec
Oct
Nov
21
3
3
6
6
3
13
5
4
2
5
16
2
3
3
13
8
10:39
02:03
01:43
09:36
18:31
16:34
10:45
18:48
23:22
10:30
18:16
09:49
21:19
16:11
16:11
10:45
22:20
Archives/
InfoWorld/
Mail/
News/
OWL/
awkscript
bin/
buckaroo
buckaroo.confused
cshrc
dickens.note
keylime.pie
owl.c
sample
sample2
src/
temp/
3. To see what aliases have been defined, use the alias command:
% alias
COPY
cp -i
DEL
rm -i
15
DIR
REN
ls
ls -lF
mv
ls -CF
4. You could improve the alias for DIR by having the output of ls fed directly into the
more program so that a directory listing with a lot of output will automatically
pause at the end of each page. To redefine an alias, just define it again:
% alias DIR ‘ls -lF | more’
To confirm that the alias is set as you desire, try this:
% alias DIR
DIR
ls -lF | more
JUST A MINUTE
If you’re just defining one command with an alias, you don’t really need
to use the quotation marks around the command argument. But what
would happen if you entered alias DIR ls -lF | more? The alias would
be set to ls -lF, and the output of the alias command would be fed to
the more program, which is quite different from what you desired.
Therefore, it’s just good form to use the quotation marks and a good habit
to get into.
Aliases are a great addition to any command shell, and with the arcane UNIX
commands, they also can be used to define full-word commands as synonyms. For
example, if you decide you’d like the simplicity of remembering only the command move to
move a file somewhere else, you could add the new alias alias move mv to your .cshrc file
if you’re using C shell or alias move=mv to your .profile if you prefer Korn shell, and the
shell would include a new command.
Task 15.4: Some Power Aliases
Because I have used the C shell for many years, I have created a variety of different
aliases to help me work efficiently. A few of the best are shown in this section.
1. To see what aliases I have defined, I can use the same command I used earlier:
% alias
cd
diff
env
from
info
library
ll
chdir !* ; setprompt
/usr/bin/diff -c -w
printenv
frm -n
ssinfo
echo “ “; echo “ “ ; echo “remember: ^J is ENTER”; tn3270 lib
ls -l
5
3
15
Hour 15
336
ls
/bin/ls -F
mail
Mail
mailq
/usr/lib/sendmail -bp
netcom echo Netcom login: taylor;rlogin netcom.com
newaliases
echo you mean newalias...
rd
readmsg $ | page
rn
/usr/local/bin/rn -d$HOME -L -M -m -e -S -/
setprompt
set prompt=”$system ($cwd:t) ! : “
ssinfo echo “connecting...” ; rlogin oasis
sunworld
echo SunWorld login: taylor;rlogin sunworld.com
Recall that each of these aliases started out in my .cshrc file surrounded by single
quote marks:
% grep alias .cshrc
alias diff
alias from
‘frm -n’
alias ll
‘ls -l’
alias ls
‘/bin/ls -F’
alias mail
Mail
alias mailq
alias netcom
‘echo Netcom login: taylor;rlogin netcom.com’
alias sunworld ‘echo SunWorld login: taylor;rlogin sunworld.com’
alias newaliases ‘echo you mean newalias...’
alias rd
alias rn
alias cd
alias env
‘printenv’
alias setprompt
# special aliases:
alias info
ssinfo
alias ssinfo
alias library
‘echo “ “; echo “ “ ; echo “remember: ^J is ENTER”;
tn3270 lib’
Also notice that the shell always keeps an alphabetically sorted list of aliases,
regardless of the order in which they were defined.
2. Most of these aliases are easy to understand. For example, the first alias, diff,
ensures that the command diff always has the default flags -c and -w. If I enter
from, I want the system to invoke frm -n; if I enter ll, I want the system to invoke
ls -l, and so on.
Some commands can cause trouble if entered, so creating an alias for each of those
commands is a good way to stay out of trouble. For example, I have an alias for
newaliases ; if I accidentally enter that command, the system gently reminds me
that I probably meant to use the newalias command:
% newaliases
you mean newalias...
3. I have created aliases for connecting to accounts on other systems. I like to name
each alias after the system to which I’m connecting (for example, netcom,
sunworld):
15
% alias netcom
echo Netcom login: taylor;rlogin netcom.com
% alias sunworld
echo SunWorld login: taylor;rlogin sunworld.com
You can’t enter alias netcom sunworld to list the netcom and sunworld
aliases because that command means to replace the alias for netcom with
the command sunworld.
CAUTION
Separating commands with a semicolon is the UNIX way of having multiple
commands on a single line, so when I enter the alias netcom, for example, it’s as if
I’d entered all these commands one after another:
echo Netcom login: taylor
rlogin netcom.com
4. Two aliases worth examining more closely are those for the cd and setprompt
commands. As you learn in a few moments, you can set your shell prompt to be
just about any characters you’d like. (Hang on for just a paragraph or two, and you
will learn all about what’s occurring in the next example!) I like to have my prompt
indicate where in the file system I’m currently working. To ensure that the prompt
is always up to date, I simply alias the cd command so that each time I change
directories, the prompt is recalculated.
% alias cd
% alias setprompt
JUST A MINUTE
The chdir command does the same thing as cd and is intended for use
within aliases as shown. So if you find chdir easier to remember than cd,
you can use it instead.
Aliases are what makes both the C shell and Korn shell such great command
interfaces. I can, and do, easily customize the set of commands and the default flags
(look at all the options I set as default values for the rn command). I even turn off some
commands that I don’t want to enter accidentally. Let your imagination run wild with aliases.
If you decide you really like one and you’re using csh, add the alias to your .cshrc so it’s
permanent (.profile if you’re using ksh). If you want to turn off an alias, you can use the
unalias command, and it’s gone until you log in again. For example, unalias netcom would
temporarily remove from the shell the netcom alias shown earlier.
7
3
15
Hour 15
338
Task 15.5: Setting Custom Prompts
Up to this point, the command prompt I’ve seen is a boring %. It turns out that
the C shell lets you set your prompt to just about any possible value, with set
prompt=”value”.
The Korn shell equivalent is even easier: PS1=”value”. Note that PS1 must be all uppercase
for this to work.
1. I’m getting tired of UNIX being so inhospitable. Fortunately, I easily can change
how it responds to me:
% set prompt=”Yes, master? “
Yes, master?
That’s more fun!
The ksh equivalent is PS1=”Yes, master? “
2. There are a lot of things you can tuck away in your prompt that can be of great
help. The first useful variable is cwd, which holds the current working directory:
Yes, master? set prompt=”In $cwd, oh master: “
In /users/taylor, oh master:
What happens if I change directories?
In /users/taylor, oh master: cd /
In /users/taylor, oh master: pwd
/
In /users/taylor, oh master:
This is not so good. Now you can see why it’s necessary to alias cd to maintain the
prompt.
3. Some special ! values can be added to the prompt definition, as shown in Table 15.2.
Table 15.2. Special values for the system prompt.
Value
Expands to
`cmd`
The results of executing cmd.
The current command number.
The value of var.
The tail (last word) of the value of var.
\!
$var
$var:t
15
Here are a few examples of other C shell prompts and what happens when you use
them:
In /, oh master: set prompt=”(\!) % “
(132) %
The ksh equivalent is PS1=”(\!) $ “.
The number in parentheses is the command number, as used by the C shell history
mechanism:
(132) % echo hi
hi
(133) % ls News
mailing.lists.usenet
(134) % !132
echo hi
hi
(135) %
usenet.1
usenet.alt
Every time I log in, I automatically set the variable system to the name of the
current computer:
(135) % set prompt=”$system (\!) % “
limbo (136) %
I like to include in my prompt the basename of the current directory, as shown in
the following example. Basename means the closest directory name, so the
basename of /usr/home/taylor is taylor, for example. Also, I replace the percent
sign with a colon, which is a bit easier to read. There is a slight problem, however;
having a : instead of % means that I have to remember I’m in C shell (or Korn
shell, as the case may be).
limbo (136) % set prompt=”$system ($cwd:t) \! : “
limbo (taylor) 137 :
4. Now I glance back at the aliases for setprompt and cd, with all these things in
mind:
limbo (taylor) 139 : alias cd
limbo (taylor) 140 : alias setprompt
You can see that the setprompt alias defines the C shell prompt as $system
($cwd:t) ! : “ , although the actual line in the .cshrc file includes the backslash
(as expected):
limbo (taylor) 141 : grep prompt= .cshrc
alias setprompt
9
3
15
Hour 15
340
Each time I change directories, I use the combined commands of the cd alias
(chdir !*) to change the current directory, and then I use setprompt to compute
the new prompt.
TIME SAVER
The !* notation in a shell alias expands to all the words you specify after
the alias word on the command line. For example, if you have an alias for
dir that is “echo !* ; ls !*” , entering dir /home actually executes echo
/home followed by ls /home.
Experiment and find a set of variables that can help you customize your UNIX
prompt. I strongly recommend that you use command numbers to familiarize
yourself with the history mechanism.
Task 15.6: Creating Simple Shell Scripts
The command-alias capability is a helpful way to cut down on entering short
commands time and again, but what if you have a series of 5 or 10 commands that
you often enter in sequence? That’s where shell scripts can help. At their simplest, shell scripts
are a series of shell commands that appears in a file in exactly the order in which they’ll be
entered. If you change the permissions of the file to add execute permission, you can enter
the name of the file as if it were just another UNIX command.
1. It’s amazing how pervasive shell scripts are in UNIX. A listing of /bin and /usr/ucb
on one system reveals that 13 and 17 commands in these files, respectively, are
actually shell scripts:
15
limbo (taylor) 33: cd /bin
limbo (bin) 34 : file * | grep script
68k:
false: executable shell script
i386:
ns32000:
pblock: executable shell script
pdp11: executable shell script
true:
u370:
u3b:
u3b10: executable shell script
u3b2:
u3b5:
vax:
limbo (bin) 38 : cd /usr/ucb
limbo (ucb) 39 : file * | grep script
msgs:
print: executable c-shell script
script: SYMMETRY i386 executable (0 @ 0) version 1
tarmail:
shell script
trman: executable shell script
uncompressdir: shell script
untarmail:
shell script
vgrind: executable c-shell script
vpq:
executable c-shell script
vpr:
vprint: executable c-shell script
vprm:
vtroff: executable c-shell script
which: executable c-shell script
zcmp:
shell script
zdiff: shell script
zmore: shell script
Shell scripts can be quite short. The script /bin/true is only one line: exit 0. The
script /bin/false contains the opposite command, and it contains only one line:
exit 1. The helpful script print is also just one line: lpr -p $* . Most of the others,
however, are too complex to explain here.
2. Instead of examining these confusing scripts, I’ll move to my own bin directory
and consider a script or two that I have there:
limbo (ucb) 42 : cd
limbo (taylor) 43 : cd bin
limbo (bin) 44 : file *
bounce.msg:
calc:
SYMMETRY i386 executable (0 @ 0) not stripped version 1
fixit: SYMMETRY i386 executable (0 @ 0) not stripped version 1
massage:
punt:
shell script
rumor.mill.sh: shell script
say.hi: ascii text
limbo (bin) 45 : cat -n punt
1 : Use /bin/sh
2
3 # Punt: punt a news article from within “rn” to yourself.
4
5 trap “/bin/rm -f /tmp/punt.$$” 0 1 9 15
6
7 [email protected]
8
9 cat - > /tmp/punt.$$
10
11 if [ “$1” != “” ] ; then
12
ADDRESS=$1
13 else
14
ADDRESS=$SENDTO
15 fi
16
17 /usr/lib/sendmail $ADDRESS < /tmp/punt.$$
18
19 echo Punted a ‘wc -l </tmp/punt.$$‘ line news article to $ADDRESS
20
21 exit 0
22
1
4
3
15
Hour 15
342
This script is intended to be part of a pipeline, and it will send a copy of the stream
of information either to the default address (SENDTO) or to a specified person ($1 is
the first argument given to the script in this case). As shown earlier in the discussion of system prompts, any text that appears in backquotes is interpreted as a
command and is executed, and the results of that command are added in its place
in the subsequent command. In this case, the echo command on line 19 computes
the number of lines in the specified file, and that number is then included in the
output, which typically looks like this: Punted a 17 line news article.
JUST A MINUTE
Notice that the very first character of this file is a colon. It turns out that the
C shell interprets scripts only if the very first character of the script is a #.
Otherwise, it lets the Bourne shell (sh) run the commands, as in this case.
3. That’s all well and interesting, but I want to create a new shell script. The first step
is to make sure that I’m creating the script in a directory that is included in my
search path (otherwise, I won’t be able to use the script as a command):
limbo (bin) 46 : pwd
/users/taylor/bin
limbo (bin) 47 : echo $PATH
.:/users/taylor/bin:/bin:/usr/bin:/usr/ucb:/usr/local:/etc:/usr/etc:
/usr/local/bin:/usr/unsup/bin (bin) 48 :
Here’s a very simple shell script that shows how shell scripts can be of assistance:
limbo (bin) 86 : cat new.script
# sample shell script
echo searching for shell scripts
pwd
echo “ “
file * | grep script | sed ‘s/:/ /’ | awk ‘{print $1}’
exit 0
This script lists the names of all files in the current directory that it identifies as
shell scripts:
limbo (bin) 88 : chmod +x new.script
limbo (bin) 89 : new.script
searching for shell scripts
/users/taylor/bin
bounce.msg
locate
new.script
punt
rumor.mill.sh
15
To confirm that the new command works, look at what file reports about this
same directory:
limbo (bin) 90 : file *
bounce.msg:
calc:
fixit: SYMMETRY i386 executable (0 @ 0) not stripped version 1
locate: shell script
massage:
new.script:
commands text
punt:
shell script
rumor.mill.sh: shell script
say.hi: ascii text
limbo (bin) 91 :
4. A more interesting script is one that can search through all the directories in my
PATH, looking for any occurrences of a specified filename:
limbo (bin) 92 : cat locate
# locate - find copies of a file
#
# this should be run by the C shell
set name=$1
foreach directory (ècho $PATH | sed ‘s/:/ /g’`)
if ( -f $directory/$name) then
ls -l $directory/$name
endif
end
The foreach loop is evaluated from the inside out. Because of the backquotes, my
PATH is echoed to sed, which removes the colons separating the directories. Then
the C shell goes through the foreach loop once for each directory in my PATH,
setting the variable directory to the subsequent value. Each time through the loop,
the -f test checks for the existence of the file: If the file exists in that directory,
ls -l lists some information about it.
Pay careful attention to the backquotes and single quotes in this script.
JUST A MINUTE
Here is locate at work:
limbo (bin)
-rwxr-xr-x
limbo (bin)
-rwxr-xr-t
limbo (bin)
93 : locate ls
1 root
32768 May 29
94 : locate vi
7 root
163840 Nov 29
95 :
1990 /bin/ls*
1990 /usr/ucb/vi*
4
3
15
Hour 15
344
It really would take an entire book (or two!) to describe fully all the ins and outs of
shell scripts. The main idea here, however, is that if you use a lot of commands
repetitively, you should make them into a command alias (if they’re short) or drop them all
into a shell script. In shell scripts, as in awk, $1 is always the first argument, $2 the second, and
so on.
Summary
This hour introduced you to many of the most powerful aspects of UNIX command shells.
Practice creating aliases and working with the history list to minimize your typing. Also, find
a prompt you like and set it in your .cshrc or .profile (for csh and ksh, respectively) so it
will be the default.
Workshop
in the next hour.
Key Terms
basename The closest directory name. For example, the basename of /usr/home/taylor
is taylor.
command number The unique number by which the shell indexes all commands. You can
place this number in your prompt using \! and use it with the history mechanism as
!command-number .
Questions
1. How do you tell the C shell that you want it to remember the last 30 commands
during a session and to remember the last 10 commands across login sessions?
2. Assume that you get the following output from entering history:
1
2
3
4
5
ls -CF
who | grep dunlaplm
wc -l < test
cat test
history
What would be the result of entering each of the following history commands?
!2
!w
!wh
echo !1
3. Some UNIX systems won’t enable you to do the following. What danger do you
see lurking in this alias?
alias who
15
who -a
5
4
3
4. Which of the following aliases do you think would be useful?
alias
alias
alias
alias
alias
alias
alias who
ls cp
copy cp -i
logout vi
vi logout
bye logout
5. Set your prompt to the following value. Remember that 33 should be replaced with
the appropriate command number each time.
#33 - I know lots about UNIX. For example:
6. Find and examine two shell scripts that are found in either the /bin or /usr/bin
directories on your system. Remember, any line beginning with a # is a comment.
In the next hour, you learn how to get even more out of your shell. You learn about shell
programming and how to create shell programs on-the-fly.
15
Hour
16
Basic Shell
Programming
In the previous hour, you learned about a few of the options available to you
when you use a command shell. These shells are how you enter commands for
UNIX. What most people don’t realize when they first start using UNIX is that
these shells are also programming languages and that you can write your own
shell programs.
Goals for This Hour
In this hour, you learn all about
■
■
■
■
■
Shell variables
Shell arithmetic
Comparison functions
Conditional expressions
Looping expressions
347
16
Hour 16
348
Because shells are really just interpreted languages, any sequence of commands you wish to
run can be placed in a file and run regularly. This is a shell program. Most UNIX experts write
their shell programs for the Bourne shell (/bin/sh) because that shell is standard on every
UNIX platform. Earlier in this book, I illustrated examples by using the C shell, because I
feel this is a better interface for the user. Because I’m programming the Bourne shell, the
command prompt is slightly different; it’s $ rather than %.
I strongly urge you to look at the shells you have available. The best shell
to use is the one that makes you the most productive.
JUST A MINUTE
Task 16.1: Shell Variables
Programming languages usually include variables, and the shell naturally does, too.
Variables are just tags to identify values that may change as a program is used. In the
shell, these variables can take a single value and are always interpreted as strings. Even numeric
values are strings to the shell.
The C shell and Korn shell both support arrays and have means of
representing numeric values.
JUST A MINUTE
You can use any string-manipulation command, such as sed or cut, to change a shell variable.
Here is an example of setting the value of a shell variable:
$ color=blue
This sets the variable color to the string blue. One can output the value of any variable with
the echo command:
$ echo $color
blue
This also indicates how to reference a shell variable: It must be preceded by the dollar sign
($). This can cause some problems. If you are using a shell variable as a prefix and want to
immediately append text, you might think this would work:
$ leaning=’anti-’
$ echo Joe is basically $leaningtaxes
16
349
The output here is just Joe is basically. The shell does not know to differentiate between
the variables $leaning and $leaningtaxes. Because there is no value assigned to $leaningtaxes,
the output is a NULL string. To solve this problem, enclose the variable in curly braces.
$ echo Joe is basically ${leaning}taxes
Joe is basically anti-taxes
If leaning is undefined, the output might not make sense. It would be Joe is basically
taxes. Fortunately, the shell provides a means to have a default value if a variable is undefined:
$ echo Joe is basically ${leaning:-pro }taxes
Joe is basically pro taxes
If leaning is undefined, the :- syntax tells the shell to use the following string, including the
space character, instead of leaving the output blank. This does not assign a new value to the
variable. If you need to use the variable repeatedly, you might want to assign a new value to
it, if it is undefined. The = character does this:
$ echo Joe is basically ${leaning=pro }taxes and ${leaning}spending.
Joe is basically pro taxes and pro spending.
The first interpretation of the variable finds it undefined, so the shell assigns ‘pro ‘ to the
variable and outputs that. The second time the variable is interpreted, it has the value ‘pro ‘.
Variables often are assigned by the read command. This assigns an individual word to a
specified variable, with the last variable in the list being assigned the remaining words.
$ read city state message
Morristown, New Jersey Hi Mom!
$ echo $city is city
Morristown, is city
$ echo $state is state
New is state
$ echo $message is message
Jersey Hi Mom! is message
As you can see, only New is assigned to state. The best way around this is to escape the space
with a backslash:
$ read city state message
Morristown, New\ Jersey Hi Mom!
$ echo $city is city
Morristown, is city
$ echo $state is state
New Jersey is state
$ echo $message is message
Hi Mom! is message
This can be a bit tricky at first.
16
Hour 16
350
The other common way to assign variables is from command-line arguments. The shell has
built-in variables to access the command line. If you’ve written a script to copy files and
named it copy-files, you might want to list all the files on the command line:
$ copy-files file1 file2 file3
The program would access these arguments as $1, $2, and $3:
cp $1 destination
cp $2 destination
cp $3 destination
The $0 variable is a special case for looking at the command name, and
command-line variables.
$*
lists all the
The standard data in any shell program is the variable. These variables can be assigned
in several ways, directly assigned, read in from a user’s typing, or by the command line.
The shell also provides means to provide some default manipulation of variables.
Task 16.2: Shell Arithmetic
Although the shell treats variables as strings, there are methods to perform some basic
mathematics on shell variables. Again, the C shell and the Korn shell provide more
extensive mathematical capabilities.
If a shell is assigned a numeric value, you can perform some basic arithmetic on the value using
the command expr. This command takes several arguments to perform arithmetic:
$ expr 1 + 1
2
Arguments must be separated by spaces, and present, for the expr command to work. If a
variable is undefined or does not have a value assigned to it (sometimes called zero length),
the result is a syntax error. Here is where the :- syntax is particularly helpful:
$ expr $undef + 1
expr: syntax error
$ expr ${undef:-0} + 1
1
Normal default values of 0 and 1 are useful. When adding 0 to a number, you get the same
number. Similarly, multiplying or dividing by 1 also doesn’t change the value of the number.
also supports subtraction, multiplication, integer division, and remainders. These are
illustrated here:
expr
$ expr 11 - 5
6
16
351
$ expr 11 ‘*’ 5
55
$ expr 11 / 5
2
$ expr 11 % 5
1
Note that I had to include the asterisk in single quotes. If I didn’t do that, the shell would
expand it to be the list of files in the current directory, and the expr program wouldn’t
understand that.
You can assign the results of the arithmetic to other variables by enclosing the command in
backquotes:
$ newvalue=‘expr ${oldvalue:-0) + 1‘
If oldvalue is assigned, it is incremented by 1. If not, newvalue is set to 1. This is useful when
looping through data for a number of iterations.
The expr command also can work with complex arithmetic. You can write an expression to
add two numbers and then multiply by a third number. Normally, you would need to worry
about operator precedence, but expr is not that sophisticated. Instead, you just group the
operations in parentheses:
$ expr $ 11 + 5 $ * 6
2
This first adds 11 and 5, then multiplies the result by 6. Because the parentheses are important
shell characters, I need to escape them with backslashes.
The expr is a very useful command for performing arithmetic in the Bourne shell.
Strings must be numbers, or there will be errors, and the results of the expr command
can be assigned to other variables.
JUST A MINUTE
The expr command is much more powerful than described here; it
includes the capability to perform logical operations and perform operations on strings. For more information, check the man page.
Task 16.3: Comparison Functions
Often, when writing a program, you may want the actions taken to be dependent
on certain values. A simple example is the rm -i command, where the –i flag tells rm
to prompt you before deleting a file. Type y, and a file is deleted. Type n, and it remains. The
shell also has similar options. These next two tasks cover how to use those options.
16
Hour 16
352
Just as expr is a powerful program for solving arithmetic expressions and performing
operations on strings, the test command can be used to perform comparisons. test will
perform comparisons on strings, as well as numeric values. Always, test will return 1 if the
condition is true and 0 if it is false. It is standard for UNIX shells to use these values as true
and false.
There are three types of operations for which test is used. There are numeric comparisons,
string comparisons, and status tests for the file system. First up are the numeric comparisons.
Because the shell treats the less-than and greater-than symbols as redirection characters, they
can’t be used within the test command. Instead, I have a series of two letter flags, as described
in Table 16.1. These flags are always placed between the two arguments:
test 3 -eq 4
This example would return false because 3 and 4 are not equal.
Table 16.1. Test operators.
Comparison
Flag
-eq
-ne
-lt
-le
-gt
-ge
Meaning
True if the numbers are equal
True if the numbers are not equal
True if the first number is less than the second number
True if the first number is less than or equal to the second number
True if the first number is greater than the second number
True if the first number is greater than or equal to the second number
You can use the result of expr, or any other command that returns a numeric value, in test.
There is also a special expression in test, -l string, that returns the length of a string. So,
you can write the following tests:
test ‘expr $value % 10‘ -eq -l $string
test ‘wc -l filename‘ -ge 10000
The first test determines if the last digit of $value (the remainder of a division by 10) is the
same as the length of $string. The second takes a count of the number of lines in a file and
is true if there are 10,000 lines or more present.
16
353
The second type of comparison is on strings. The first two are unary, which means they apply
to only one string:
test -z $string
test -n $string
The first test is true if the string is of zero length. If the string is undefined, this is true, too.
The second is true if the string has some content.
The next two tests compare strings with each other. The simple equals sign and the
exclamation point (commonly used to switch between true and false in UNIX) are used for
these comparisons:
test alphabet = Alphabet
test alphabet != Alphabet
The first is false; the second is true.
When comparing string variables, you may see something like
test X$string1 = X$string2
JUST A MINUTE
The presence of the X prevents a null string from confusing test. If
string1 is null and string2 is string , you’d expand to
test X = Xstring
Without the X, the test would be expanded to
test = string
This is a syntax error. The other option is to enclose the string in double
quotes:
test “$string1” = “$string2”
That expands to this
test “” = “string”
The final test operators work on the file system. They are single flags, listed in Table 16.2,
followed by a path.
Table 16.2. File system unary flags.
Option
Meaning
-G
True if the file exists and is owned by the same group as the process.
True if the file exists and points to another file (symbolic link).
-L
continues
16
Hour 16
354
Option
Meaning
-O
True if the file exists and is owned by the same user as the process.
True if the file exists and is a file used for communications between
programs (socket).
True if the file exists and is a symbol identifying a physical device used for
input and output in large chunks of data, such as a hard disk, (block special
device).
True if the file exists and is a symbol identifying a physical device used for
input and output in single characters, such as a terminal (character special
device).
True if the file is a directory.
True if the file exists.
True if the file exists and is a regular file.
True if the file exists and runs in a specific group.
True if the file exists and is set to remain in memory after execution. This
makes a program a faster starter, at the cost of overall system performance.
True if the file exists and is a named pipe.
True if the file exists and is readable.
True if the file exists and has data.
True if the file exists and runs as a specific user.
True if the file exists and is writable.
True if the file exists and is executable.
-S
-b
-c
-d
-e
-f
-g
-k
-p
-r
-s
-u
-w
-x
A sample test would be:
test -d $HOME/bin
This checks to see if you have a directory named bin in your home directory. The most
common flags you see in shell programs are the -f flag and the -d flag. The others are used
only in unusual situations.
The file system also has three binary comparisons. The -ef test determines whether the two
files are the same. (When you create a link between files, this is true.) The -nt flag is true if
the first file is newer than the second, and the -ot flag is true if the first file is older than the
second. You might see a test in a looping statement like:
test file1 -ot file2
16
355
This test compares the two files and is true if file1 is older than file2. If you are waiting for
data to appear in file1, you might use this test to cause a shell program to wait for the first
file to appear.
You can negate test commands with the exclamation point or combined with -a for and and
-o for or. You can make arbitrarily long conditions, at the cost of readability:
test $var -eq 0 -a ! -e file
This checks to see if the value of $var is zero and if file exists.
The test command also has a second form. Instead of explicitly calling test, the condition
is surrounded by square brackets:
[ -f file ]
Doing this makes shell programs more readable.
One of the most-used commands in shell programming is the test command. It is
essential to understanding the next two tasks, conditional expressions and loops.
Task 16.4: Conditional Expressions
Sometimes, when writing a program, you want to perform an action only when
another action returns the value true. Shell programming enables you to do this by
way of the if command, the case command, and two special command separators.
The if command is the most commonly seen conditional command. It takes the form:
if
command-block
then
command-block
fi
A command-block is a sequence of one or more shell commands. The first command-block
is always executed. The return value of the last statement executed is used to determine if the
second block is executed. The most commonly used command at the end of the first
command-block is the test command.
if
[ -f $file ]
then
echo $file is a regular file
fi
16
Hour 16
356
This if statement notifies the user that a file is a regular file. If the file is not a regular file (such
as a directory), you don’t see output.
Sometimes, you may want output regardless of the situation. In the preceding case, you may
be interested in the status of the file even if it is not a regular file. The if command can expand
with the else keyword to provide that second option.
if
[ -f $file ]
then
echo $file is a regular file
else
echo $file is not a regular file
fi
This statement provides output regardless of the status of the file.
For these simple tests and output, the shell provides a second, quicker means of executing the
if statement. If the two commands are joined by && , the second command is executed if the
first command is true. If the commands are joined by ||, the second command is executed
if the first is false. The preceding command, therefore, would look like:
[ -f $file ] && echo $file is a regular file
[ -f $file ] || echo $file is not a regular file
This shorthand is very useful but can be confusing for a novice. If you accidentally place a
space between the characters, you have a wildly different command; the & will run the first
command at the same time as the echo, and the | will pipe the output of the test (none) to
the echo.
If you want even more information, your if statement can have more than two options. You
need multiple tests and the elif keyword:
if
[ -f
then
echo
elif
[ -d
then
echo
else
echo
fi
$file ]
$file is a regular file
$file ]
$file is a directory
$file is not a regular file or a directory.
This command first tests to see whether the file is a regular file; if not, it checks to see whether
it is a directory, and if it is neither, it gives output. You can expand any if statement with an
unlimited number of elif branches.
16
357
At some point, though, the code will become confusing. When you have many possible
branches, you should use the case command. The syntax is more complicated than for if:
case string in
pattern) command-block ;;
pattern) command-block ;;
...
esac
If you were looking for possible values for a variable, you could use case:
echo What do you want:
read var remainder
case $var in
house) echo The price must be very high;;
car) echo The price must be high;;
popsicle) echo The price must be low;;
*) echo I do not know the price;;
esac
This case statement follows an input request and gives the user a rough idea of the price. A
case list can contain any number of items.
The pattern-matching algorithms used are for wildcards.
There are two basic conditional expressions and a third shortcut. You can test a
condition and perform alternative actions by using if statements and their shortcuts.
Or, you can compare strings and perform any number of actions by using the case
statement.
Task 16.5: Looping Expressions
If you want to run the same set of commands many times instead of writing them
out once for each time, you are better off using looping commands. There are two
types of loops, the determinate loop and the indeterminate loop.
A determinate loop is one where you know exactly how many times you want to execute the
commands before you enter the loop. Stepping through a list of files is a good example; you
may not know the exact number of files, but once you do, you can start the loop for those
files.
An indeterminate loop is one where you need to keep executing a command-block until a
condition is no longer true. You might be either waiting for something or performing a series
of modifications to reach a goal.
16
Hour 16
358
The usual command for a determinate loop is the for command. It has the following syntax:
for var in list
do
command-block
done
You can build any list you like. It could be a sequence of numbers or the output of a command.
Earlier, I mentioned looping through a list of files. This is performed with the following loop:
for var in ‘ls‘
do
if
[ -f $var ]
then
echo $var is a regular file
fi
done
This provides a list of all the regular files. You should note that the variable var can be used
inside the for loop. When you are stepping through a list of files, this can be advantageous;
you have the name of the file provided to you by the variable var.
A nice trick that can be performed in a shell program is to step through the list of commandline arguments. The for loop provides a neat mechanism; if the ‘in list’ part is omitted
from the command, the for loop steps through the list of command-line arguments.
j=0
for i
do
j=‘expr $j + 1‘
echo $i is argument $j
done
This command steps through the command-line arguments and identifies where they are in
the order of arguments.
In both cases, when you enter the for loop, you know how many times you need to run the
loop. If you look at the case where you are waiting for something to happen, though, you need
to use a different loop. The while loop is the solution for this problem.
In Task 16.3, I mentioned the case where you might want to wait on the arrival of a file. This
echoes a real-world situation I recently faced. We were processing a daemon’s log file, but we
did not know exactly when it would be placed in our directory. We tried to set up the job to
run after the file arrived, but this still ran into problems.
16
359
Using the while loop, we solved the problem. At the end of the execution of our script, we
created a checkpoint file. At the beginning, if the checkpoint file was newer than the log file,
we’d wait. Programmatically, that is:
while
[ checkpoint -nt logfile ]
do
sleep 60
done
This program would wait one minute between checks. If the new logfile had not been
written, the program would go back to sleep for a minute.
You can use while loops also in a determinate manner. In the case where you are not
concerned with a variable’s value but know a count of times to run a command-block, you
can use a counter to increment through the number:
i=0
while
[ $i -lt 100 ]
do
i=‘expr “$i” + 1‘
commands
done
This is certainly easier than listing 100 items in a list!
The shell provides two convenient mechanisms for running a group of commands
repeatedly. These loop commands are useful from both the command line and a
program.
Summary
In this hour, you just skimmed the basics of shell programming. You were introduced to the
control structures of the shell and two important commands. There is a lot more you can learn
about shell programming; Teach Yourself Shell Programming (Sams Publishing) is one place
you can look. If you are interested in the C shell, Teach Yourself the UNIX C Shell
(Sams Publishing) is another resource.
Workshop
in the next hour.
16
Hour 16
360
Key Terms
command-block A list of one or more shell commands that are grouped in a conditional
or looping statement.
conditional expression This is an expression that returns either true or false.
determinate loop A loop where the number of times the loop is run can be known before
starting the loop.
expression This is a command that returns a value.
indeterminate loop A loop where the number of times the loop is run is not known before
starting the loop.
loop This is a sequence of commands that is repeatedly executed while a condition is true.
variables These are names to label data that may change during the execution of a program.
zero-length variable
A variable that does not have a value assigned to it.
Questions
1. How would you read in an address in a shell program? How would you read in a
name?
2. If you read in the number of people who read a newspaper and the number of
people who subscribe to a particular paper, how would you determine the ratio of
subscribers to readers?
3. How do you know if a file has data?
4. How do you wait for data to be placed in a file?
In the next hour, you are introduced to managing processes in UNIX. You learn how to start
a background job, how to switch between foreground and background, and how to terminate
a command.
16
Job Control
Hour
17
Job Control
In this hour you will learn about how UNIX handles jobs, and how you can
manipulate them. Commands you will learn include jobs and ps, to see what
processes are running; fg and bg, to move jobs back and forth between the
foreground and background; and kill, to terminate jobs that you no longer
want around.
Goals for This Hour
■ About job control in the shell: stopping jobs
■ How to put jobs in the background and bring them back to the
foreground
■ How to find out what tasks are running by using jobs and ps
■ How to terminate errant processes by using kill
Throughout this book, I’ve indicated that my focus is on the most important
and valuable flags and options for the commands covered. That’s all well and
good, but how do you find out about the other alternatives that might actually
work better for your use?
361
17
Hour 17
362
This hour presents an explanation of a UNIX philosophical puzzle: What is a running
program? To learn the answer, you are introduced to ps and jobs, for controlling processes;
fg and bg, to move your own processes back and forth between the foreground and
background; and the quasi-omnipotent kill command, for stopping programs in their
proverbial tracks.
Task 17.1: Job Control in the Shell: Stopping Jobs
Whether you’re requesting a man page, listing files with ls, starting vi, or running
just about any UNIX command, you’re starting one or more processes. In UNIX,
any program that’s running is a process. You can have multiple processes running at once. The
pipeline ls -l | sort | more invokes three processes: ls, sort, and more. Processes in both
the C and Korn shells are also known as jobs, and the program that you’re running is known
as the current job.
Any job or process can have a variety of different states, with “running” being the most typical
state. In both shells, you can stop a job by pressing ^z. To restart it, enter fg when you are
ready.
1. Earlier I was perusing the man page entry for sort. I had reached the bottom of the
first screen:
% man sort
SORT(1)
SORT(1)
NAME
sort - sort or merge files
SYNOPSIS
sort [ -mubdfinrtx ] [ +pos1 [ -pos2 ] ] ... [ -o name ] [
-T directory ] [ name ] ...
DESCRIPTION
Sort sorts lines of all the named files together and writes
the result on the standard output. The name ‘-’ means the
standard input. If no input files are named, the standard
input is sorted.
The default sort key is an entire line. Default ordering is
lexicographic by bytes in machine collating sequence. The
b
Ignore leading blanks (spaces and tabs) in field com--More-- _
17
Job Control
363
I’d like to try using the -b flag mentioned at the bottom of this screen, but I want
to read the rest of the man page, too. Instead of typing q to quit, then starting the
man program again later, I can stop the program. I press ^z, and see this:
b
--More-Stopped
%
Ignore leading blanks (spaces and tabs) in field com-
At this point, I can do whatever I’d like:
% ls
1
1
1
1
1
17
-s | sort -b | head -4
Archives/
InfoWorld/
Mail/
News/
OWL/
2. I can resume at any time. I enter fg, the program reminds me where I was, and man
(which is actually the more program invoked by man) returns to its prompt:
% fg
man sort
--More-- _
%
3. Screen-oriented programs are even smarter about stopping and starting jobs. For
example, -vi refreshes the entire screen when you return from it having been
stopped. If I were in vi working on the dickens.note file, the screen would look
like this:
Preface
care and interest.
Hour 17
364
Pressing ^z would result in this:
Stopped
% _
I can check to see if someone is logged in and then return to vi with the fg
command.
% who | grep marv
% fg
Preface
care and interest.
17
Job Control
365
There are many aspects to processes and jobs in UNIX, particularly regarding the level
of control offered by the shell. The rest of this hour explains how to exploit these
capabilities to make your work easier and faster.
Task 17.2: Foreground/Background and UNIX
Programs
Now that you know how to suspend programs in their tracks, it’s time to learn how
to have them keep running in the background (by using the bg command) while
you’re doing something else and how to have programs instantly go into the background (by
using the & notation).
In the first hour, you learned that one of the distinguishing characteristics of UNIX is that
it’s a true multitasking operating system. It is capable of running hundreds of programs at
the same time. The best part is that you’re not limited to just one process! If you want to save
a couple man pages to a file, for example, you can run those processes in the background while
you are working on something else.
Once a job is stopped, you can enter fg to start it up again as the program you’re working
with. (The fg command takes its name from foreground, which refers to the program that your
display and keyboard are working with.) If the process will continue without any output to
the screen and without any requirement for input, you can use bg to move it into the
background, where it runs until it is done. If the program needs to write to the screen or read
from the keyboard, the system will stop its execution automatically and inform you. You then
can use fg to bring the program into the foreground to continue running.
If you find that background jobs are just writing information to your
screen, try the stty tostop command to fix the problem.
TIME SAVER
You can use job control also to start up a couple programs and then use the fg command, with
the job ID as an argument to start the one you want to work with. Not entering the job ID
will bring the most recently stopped job back to the foreground. If your system takes a long
time to start up big applications (such as emacs or vi), this could save you lots of time.
JUST A MINUTE
Although a job may be stopped, it still consumes resources, so you should
be careful not to have too many stopped programs around, in deference
to the other users of your machine.
17
Hour 17
366
A different strategy is to start a program in the background, letting UNIX manage it. If the
program needs some input or output, it stops, just like processes you’ve put into the
background with bg after they’ve already started running. To have a program (or pipeline!)
automatically start in the background, simply type an & at the end of the command line.
1. Here’s an example of a command that processes files without needing any input or
offering any output:
% awk -F: ‘{print $1" = “$5}’ < /etc/passwd | \
awk -F, ‘{print $1}’| \
awk ‘{ if (NF > 2) print $0 }’ | \
sort > who.is.who
After about 20 seconds, the % prompt returns; it takes that long to feed the password file through the three-part awk filter, sort the entire output, and save it to the
file who.is.who.
CAUTION
When you’re working with long commands, it’s useful to know that you
always can move to the next line—even in the middle of entering something—by ending the current line with a single backslash. Note that the
backslash must be the very last character on the line!
With this new file, I easily can look up an account to see the full name of that user:
% alias lookup ‘grep -i \!* who.is.who’
% who | head
root
console Dec 6 18:02
maritanj ttyAa
Dec 8 21:20
efb
ttyAb
Dec 8 12:12
wifey
ttyAc
Dec 8 19:41
phamtu
ttyAe
Dec 8 21:14
curts
ttyAf
Dec 8 21:14
seifert ttyAg
Dec 8 21:11
taylor
ttyAh
Dec 8 21:09
halcyon ttyAi
Dec 8 18:34
jamilrr ttyAj
Dec 8 20:25
Broken pipe
% lookup maritanj
maritanj = Jorge Maritan
% lookup efb
efb = Edward F. Billiard
%
17
Job Control
367
2. To have this process run in the background, I can stop the process immediately
after I start it, by using ^z:
% !awk
awk -F: ‘{print $1" = “$5}’ < /etc/passwd | awk -F, ‘{print $1}’
➥| awk ‘{ if (NF > 2) print $0 }’ | sort > who.is.who
Stopped
%
Notice that the command I repeated using the history mechanism was
listed as being all on a single line!
JUST A MINUTE
At this point, bg will continue the program, running it in the background:
% bg
[1]
awk -F: {print $1" = “$5} < /etc/passwd | awk -F, {print $1}
➥| awk { if (NF > 2) print $0 } | sort > who.is.who &
%
The number in square brackets is this job’s control number in the shell. In a
moment, you learn why this is a handy number to note.
On some systems a completed background job will notify you immediately that it’s
done, but on most systems, after a completed background job has finished running,
it waits until you press Return to get a new system prompt before it lets you know.
After about 30 or 40 seconds, I press Return and see this:
%
[1]
Done
awk -F:
{print $1" = “$5} < /etc/passwd | awk -F,
{print $1} | awk { if (NF > 2) print $0 } | sort > who.is.who
%
3. Alternatively, a better strategy for moving a program into the background is to
move the process to the background automatically by adding an & to the very end:
% !awk &
➥| awk ‘{ if (NF > 2) print $0 }’ | sort > ! who.is.who &
[1] 27556 27557 27558 27559
%
This is more interesting. This command is shown with a control number of 1, but
the four numbers listed after that are the actual process ID numbers of each piece
of the pipeline: 27556 is the first awk process, 27557 is the second awk process, 27558
is the third awk process, and 27559 is the sort program.
Again, when complete, pressing Return lets me know:
%
[1]
Done
awk -F: {print $1" = “$5} < /etc/passwd |
➥awk -F, {print $1} | awk { if (NF > 2) print $0 } | sort > who.is.who
%
17
Hour 17
368
4. What happens if I try to automatically move to the background a program that has
input or output?
% vi &
[1] 28258
%
This looks fine. Pressing Return indicates otherwise, though:
%
[1]
%
+ Stopped (tty output) vi
You can see that this program has stopped because of some information (output)
that it wants to display. If the program expected input, the message would be
Stopped (tty input) program name .
I can use fg to bring this program into the foreground and work with it, or even
just to quit vi.
Because so much of the UNIX design focuses on running streams of data through
filters and saving the output to a file, there are a number of commands that you could
be running in the background, freeing you up to do other work in the meantime. Remember
also that you can put in the background jobs that take a fair amount of processing time and
then display information on the screen. When it’s time to write something to the screen, the
program will stop automatically until you enter fg to pull it into the foreground again.
Task 17.3: Finding Out What Tasks Are Running
There are two ways to keep tabs on what programs are flying around in the UNIX
operating system. The easier way, jobs, shows what processes you’ve stopped and
moved into the background in the shell. Enter jobs, and csh (or ksh) tells you what programs,
if any, are stopped or running.
The alternative is a complex command called ps, which shows the processor status for the
entire computer. The processor is another name for the computer itself. Fortunately, without
any arguments, it shows the active or stopped programs associated with your terminal only.
The ps program actually has more flags even than ls, I think. The vast majority of them,
however, are never going to be of value to you or any normal UNIX user. Worse, the flags
are very different between BSD systems and System V. The ones that are most helpful are
summarized in Table 17.1.
Table 17.1. Useful flags to the ps command, BSD-style.
Flag
Meaning
-a
Shows all processes associated with terminals attached to the system.
Shows all interesting processes on the system (that is, all processes other than
those required by the operating system).
-g
17
Job Control
Flag
Meaning
-l
Gives the long listing format for each line.
Lists only processes associated with the specified ttyxx.
Produces user-oriented output.
Uses wide output format. If repeated (-ww), it will show as much of each
command as possible.
Shows all processes in the system.
-t xx
-u
-w
-x
The -a, -g, and -x flags each affect how much information is displayed by ps. To use either
the -g or -x command, you also must use the -a command. On most machines, -ax yields
considerably more output than -ag. The most commonly used flags (and flag combinations)
are -u, to have only your processes listed in a friendly format; -aux, to see everything on the
machine (you almost always want to pipe this to grep or more, lest you be overrun with
hundreds of lines of information); and -wtxx, to show all the processes associated with ttyxx,
in wide format.
JUST A MINUTE
The ps program varies from System V to Berkeley UNIX more than any
other command. Fortunately, the two or three most common flags are
similar across the two systems. To explore more about the ps command on
your system, you should start by reading the man page.
1. To begin, I’m going to start vi in the background:
% vi dickens.note &
[1] 4352
%
I’ll start that awk job again, too:
% !awk
➥| awk ‘{ if (NF > 2) print $0 }’ | sort > ! who.is.who &
[2] 4532 4534 4536 4537
%
The jobs command will show what processes I have running:
% jobs
[1] + Stopped (tty output) vi dickens.note
[2] - Running
awk -F: {print $1" = “$5} < /etc/passwd | awk F,
{print $1} | awk { if (NF > 2) print $0 } | sort > who.is.who
%
369
17
Hour 17
370
2. Now that you know the job numbers (the numbers in square brackets here), you
easily can move specific jobs into the foreground or the background by specifying
the job number prefixed by %. To show what I mean, I’ll put a couple more vi jobs
in the background:
% vi buckaroo.confused &
[2] 13056
% vi awkscript csh.man cheryl mbox &
[3] 13144
%
Now I’ll use the jobs command to see what’s running:
% jobs
[1]
Stopped (tty output) vi dickens.note
[2] - Stopped (tty output) vi buckaroo.confused
[3] + Stopped (tty output) vi awkscript csh.man cheryl mbox
%
Notice that the awk job finished.
JUST A MINUTE
To edit the buckaroo.confused note, I need only to enter fg %2 to pull the file into
the foreground. To terminate all these processes (something you learn more about
later in this hour), I can use the kill command:
% kill %1 %2 %3
%
Nothing happened. Or did it? Pressing Return reveals what occurred in the
operating system:
%
[3]
[2]
[1]
%
- Done
- Done
+ Done
vi awkscript csh.man cheryl mbox
vi buckaroo.confused
vi dickens.note
3. Restart the awk command with !awk. Contrast the output of jobs with the output
of the Berkeley (BSD) ps command:
% ps
PID
4352
4532
4534
4536
4537
4579
%
17
TT
Ah
Ah
Ah
Ah
Ah
Ah
STAT
T
R
R
S
S
R
TIME
0:00
0:03
0:02
0:01
0:00
0:00
COMMAND
vi dickens.note
awk - : {print $1"
awk - , {print $1}
- k { if (NF > 2) print $0 } (awk)
sort
ps
Job Control
371
You can see here that there really are four unique processes running for that
pipeline: three awk processes and one sort process. In addition, vi and ps are listed
as running. Note that my login shell (csh) isn’t in this listing.
Figure 17.1 explains each field, and Table 17.2 lists possible values for the STAT
program status column.
Figure 17.1.
The ps default process
output.
Table 17.2. Possible process status values.
Value
Meaning
R
Running
Sleeping (20 seconds or less)
Idle (sleeping more than 20 seconds)
Stopped
Zombie process
S
I
T
Z
There are other process states, but they rarely show up for most users. A zombie
process is one that has ended but hasn’t freed up its resources. Usually, it takes a
second or two for the system to completely recover all memory used by a program.
Sometimes, zombies are stuck in the process table for one reason or other. UNIX
folk refer to this as a wedged process, which stays around until the system is
rebooted. Sometimes it’s listed as <defunct> in process listings. Any process that is
preceded by a sleep command is noted as sleeping.
4. Adding some flags can change the output of ps quite dramatically:
% ps -x
PID TT
4352 Ah
6171 Ah
6172 Ah
6173 Ah
6174 Ah
6177 Ah
19189 Ah
19649 Ah
%
STAT
T
R
R
S
S
R
S
I
TIME
0:00
0:02
0:01
0:01
0:00
0:00
0:06
0:02
COMMAND
vi dickens.note
awk - : {print $1"
awk - , {print $1}
- k { if (NF > 2) print $0 } (awk)
sort
ps -x
-csh (csh)
newmail
17
Hour 17
372
Two new processes show up here: -csh (the shell) which is—finally—my login
shell, and newmail, a program that automatically starts up in the background when
I log in to the system (it’s located at the end of my .login).
JUST A MINUTE
The shell process is shown with a leading dash to indicate that it’s a login
shell. Any other copies of csh that I run won’t have that leading dash.
That’s one way the C shell knows not to read through the .login file every
time it’s run.
5. To see more about what’s happening, I add yet another flag, -u, to expand the
output on the display:
% ps -xu
USER
PID
taylor
7011
taylor
7012
taylor
7013
print
taylor
19189
taylor
7014
taylor
7022
taylor
4352
taylor
19649
%
%CPU %MEM
10.4 0.2
6.3 0.1
5.9 0.1
1.1
1.0
0.1
0.0
0.0
0.2
0.1
0.2
0.3
0.1
SZ
184
160
160
RSS
100
92
92
TT
Ah
Ah
Ah
STAT ENG
R
6
S
R
3
TIME
0:02
0:01
0:01
COMMAND
awk - : {print $1"
awk - , {print $1}
- k { if (NF > 2)
256
316
180
452
124
148
64
116
168
60
Ah
Ah
Ah
Ah
Ah
S
S
R
T
I
0:07
0:00
0:00
0:00
0:02
-csh (csh)
sort
ps -xu
vi dickens.note
newmail
0
Figure 17.2 explains these fields.
Figure 17.2.
The -u user-oriented
output of ps.
6. I won’t show the output from the -aux flags, but you should look at the number of
lines produced by both the -ag and -ax flags:
% ps -ag
377
% ps -ag
PID TT
1403 co
2200 p3
17
| wc -l
| head
STAT TIME COMMAND
IW
0:01 -csh (csh)
IW
0:18 server
Job Control
6076 p6 I
6082 p6 I
25341 p8 IW
681 pa IW
10994 pa IW
11794 pa IW
13861 pa I
Broken pipe
%
0:13
0:11
0:06
0:05
2:10
0:12
0:56
rlogin sage -l hirschna
rlogin sage -l hirschna
-tcsh (tcsh)
-tcsh (tcsh)
ghostview pop5.ps
pwlookup
gs
You can see here that each process is owned by a specific terminal but that these
processes are all idle (that is, they’ve been sleeping for more than 20 seconds). This
probably means that these users have turned away for a little while. Look back at
the output generated by ps -xu, and you will see that newmail is also idle. That’s
because the program runs in a loop: It sleeps for five minutes, checks for new mail,
goes back to sleep again, and so on. Processes that have the W after the I in the
status column are processes that have been moved out of main memory and are
swapped out to disk. This is not a problem, and the users might not even realize
anything has happened; the only symptom of this is that when the users wake up
their programs, the programs will take an additional second or two to return.
What is the output from ps -ax?
% ps -ax | wc -l
765
% ps -ax | head
PID TT STAT TIME
0 ? D
8:58
1 ? S
14:45
2 ? D
20:43
27 ? I
0:00
59 ? S
6:36
70 ? I
0:02
74 ? IW
0:00
75 ? IW
0:00
76 ? IW
0:00
Broken pipe
%
373
COMMAND
swapper
(init)
pagedaemon
rpc.rquotad
/etc/syslogd -m480
/etc/portmap
(biod)
(biod)
(biod)
These are some of the “guts” of the UNIX operating system. Notice that none of
these processes are actually associated with a terminal. Also notice that some of
these processes have incredibly low process ID numbers! Any one-digit process ID
is a program that is a part of the core UNIX system and must be running for UNIX
to be alive. Any two-digit process is also started by the system itself but is probably
optional. The D status for some of these processes indicates that they’re waiting for
disk resources of some sort. Finally, note how much time these processes have
taken. I venture that you will never have a process that takes 20 minutes of CPU
time—ever!
17
Hour 17
374
7. On a Sun workstation, the output of the ps commands is a bit different:
% ps
PID
8172
8182
8186
%
TT
qb
qb
qb
STAT
S
T
R
TIME
0:00
0:00
0:00
COMMAND
-csh (csh)
vi
ps
In many ways, though, these different workstations have very similar output from
the ps commands. For example, compare this Sequent output from ps -xu to the
ps -xu output on the Sun that I already showed:
% ps -xu
USER
taylor
taylor
taylor
taylor
%
PID %CPU %MEM
8191 7.7 0.4
8182 0.0 0.4
8172 0.0 0.3
8180 0.0 0.1
SZ
284
140
68
52
RSS
536
432
400
144
TT
qb
qb
qb
qb
STAT
R
T
S
S
START
19:16
19:16
19:16
19:16
TIME COMMAND
0:00 ps -xu
0:00 vi
0:00 -csh (csh)
0:00 newmail
The ENG column of the previous examples is replaced by a START column on the Sun
workstation. The numbers in the ENG column indicate the exact time that the
processes were started on the computer.
UNIX works with processes. Your login shell, the edit session you run, and even the
ls program listing your files are all processes in the operating system. This means that
you can work with processes. You can stop programs temporarily to do something else, restart
them as you choose, and even look at all the programs you’re running at any time, including
otherwise hidden processes such as your login shell itself.
Task 17.4: Terminating Processes with kill
Now that you know how to create multiple processes, tuck some into the background, and find stray processes, you need some way to permanently stop them from
running, as needed. The command to accomplish this in UNIX is kill. For the most part,
to use kill, you specify the process ID numbers of those programs you want to terminate.
Both the C shell and Korn shell have a convenient shorthand that you’ve already seen: the
percent–job-number notation.
There are a variety of different signals that the kill command can send to a process. To specify
a job control action, you need to specify to kill one of a variety of different signals.
Table 17.3 lists signals you can use with kill.
Table 17.3. Some signals to use with kill.
17
Number
Name
Meaning
1
2
SIGHUP
Hang up
Interrupt
SIGINT
Job Control
Number
Name
Meaning
9
15
SIGKILL
kill
SIGTERM
(cannot be caught or ignored)
Software termination signal from kill
There are over 30 different signals that UNIX knows about, but Table 17.3 lists the ones that
are most helpful. The SIGHUP signal is what’s sent to every process you are running just before
you hang up (log out of the system). SIGINT is the signal sent when you press ^c; many
programs respond in specific ways when this signal is received. SIGKILL is “The Terminator”
of the UNIX signals: Programs cannot ignore it and cannot process it. The process is
terminated immediately, without even a chance to clean up after itself. SIGTERM is the more
graceful alternative: It requests an immediate termination of the program, but it allows the
program an opportunity to remove temporary files it might have created.
By default, kill sends a SIGTERM to the processes specified. You can specify other signals,
however, by using either the number or the name of the signal (minus the SIG prefix, that is).
On many systems, you also can specify the -l flag to kill to see what signals are available.
CAUTION
The kill command should be used with caution. It can get you into a lot
of trouble. For example, do you want to log out rather suddenly? To do
that, find the process ID of your login shell and terminate it. Learn to use
kill, but learn to use it cautiously.
1. The simplest way to use the kill command is from the shell. First, start a job in
the background:
% vi &
[1] 6016
%
I can terminate this process now by using either kill %1 or kill 6016, but if I try
both of them, the second will fail because the first already will have terminated the
process:
% kill %1
% kill 6016
6016: No such process
[1]
Done
%
375
vi
Just as if I had dropped a process into the background and it instantly stopped
because it needed to produce output, the kill process also had no feedback and
17
Hour 17
376
took a second or two to occur. In the interim, I entered the second kill command,
which then output the error message No such process. Following that, I get an
indication from the shell itself that the job ended.
2. Using the ps command, I can find that pesky newmail program that’s always
running in the background:
% ps -ux | grep newmail
taylor
6899
0.1 0.1
taylor
25817
0.0 0.1
%
52
124
28 Av S
60 Av I
0:00 grep newmail
0:01 newmail
I want to send that process a hang-up signal (SIGHUP):
% kill -HUP 25817
% !ps
ps -ux | grep newmail
taylor
7220
0.0 0.1
%
52
28 Av S
0:00 grep newmail
Because the newmail program isn’t in this listing, I can’t conclude that the SIGHUP
signal stopped newmail.
JUST A MINUTE
Because kill tells you if a process cannot be found, the typical UNIX
solution to finding out if the command worked is to enter !! immediately
to repeat the kill command a second time. If kill worked, you see No
such process .
3. Some processes are pesky and can resist the less powerful signals SIGTERM and
SIGHUP. (In UNIX, this is called “catching” a signal. In some processes, you need to
send and catch signals to perform certain actions.) That’s when you need to use
what I call “The Big Guns,” or SIGKILL. You see this referred to sometimes as the
terminate-with-extreme-prejudice command; the format is kill -9 processID, and
it’s not for the faint of heart!
I strongly recommend that you just let kill send the SIGTERM signal and see if that
does the job. If it doesn’t, try SIGHUP, and if that also fails, use SIGKILL as a last
resort.
4. What happens if you try to use kill on jobs that aren’t yours? Fortunately, it
doesn’t work:
% ps -aux | head -5
USER
PID %CPU %MEM
SZ
news
7460 97.7 0.4 336
➥ /tmp/nnsubj6735a
phaedrus 8693 18.1 1.1 1260
root
8741 14.4 0.4 416
root
8696 13.9 0.4 416
17
RSS TT STAT ENG
252 ? R N
4
TIME COMMAND
4:33 sort –u
720 rm S
252 ? R
252 ? S
0:03 nn
0:03 nntpd
0:03 nntpd
9
Job Control
377
Broken pipe
% kill 7460
7460: Not owner
%
5. Finally, if you forget and leave stopped jobs in the background and try to log out,
here’s what happens:
% logout
There are stopped jobs.
%
You must either use fg to bring each job into the foreground and terminate them
normally, or use kill to terminate each of the jobs, then log out.
In this task, you have been introduced to the kill command and some of the signals
associated with it.
Summary
Although the file is the underlying unit in the UNIX file system, including all directories, the
most fundamental piece of UNIX is the process. In this hour, you learned how to have
background processes, how to stop and restart processes, and how to use kill to quit any
errant program—running or not.
Workshop
in the next hour.
Key Terms
control number A unique number that the C shell assigns to each background job for easy
reference and for using with other commands, such as fg and kill.
current job The job that is currently running on the terminal and keyboard (it’s the
program you’re actually running and working within).
foreground job A synonym for current job.
errant process A process that is not performing the job you expected it to perform.
job A synonym for process.
kill
Terminate a process.
17
Hour 17
378
login shell The shell process that started when you logged in to the system. This is usually
where you’re working when you’re logged in to UNIX.
process A program stopped or running within the UNIX operating system. Also known
as a job.
signals Special messages that can be sent to stopped or running processes.
stop a job Stop the running program without terminating it.
wedged process A process that is stuck in memory and can’t free up its resources even
though it has ceased running. This is rare, but annoying.
zombie
A terminated process that has not been cleaned up by the parent process.
Questions
1. Start a program, such as vi, and use ^z to stop it. Now terminate the process using
kill.
2. Start vi again, stop it, and put it in the background. Work on something else, and
then return vi to the foreground.
3. Use ps to check the status of processes to see what processes you have running that
aren’t shown on jobs. Why might ps and jobs list different processes?
The next hour focuses on the many facets of printing and generating hard copy on the UNIX
system. It’s not as easy as you might think, so stay tuned!
17
Hour
18
Printing in the UNIX
Environment
One of the greatest shortcomings of UNIX is printing. Generating printouts
is a sufficiently common task that it should be fairly easy to accomplish.
However, in this one area of UNIX, there has been continual conflict between
the System V and BSD groups, to the detriment of all.
This hour focuses on some of the most common UNIX commands for working
with printers. It is a primer on how to find out what printers are hooked up to
your system, how to send output to a printer, how to check that your print
requests are in the queue for printing, and how to remove your print requests
from the queue if you change your mind for any reason.
Goals for This Hour
■
■
■
■
379
Find local printers with printers
Send a print job to a printer with lpr or lp
Format print jobs with pr
Work with the print queue by using lpq, lprm
18
Hour 18
380
Various techniques can minimize the complexity of printing in UNIX, the best of which is
to create an alias called print that has all the default configuration information you want. If
you define PRINTER as an environment variable, most of the UNIX print utilities will default
to the printer you specify as the value of the PRINTER environment variable, for example, when
searching print queues for jobs. The queue, or list, is where all print jobs are placed for
processing by the specific printer.
JUST A MINUTE
The differing “philosophies” of BSD and System V have caused problems
in the area of printing. In a nutshell, because UNIX systems are always
networked (that is, hooked together with high-speed data-communications
lines), the most valuable feature of a printing tool would be allowing the
user to choose to print on any of the many printers attached. For this to
work, each machine with an attached printer must be listening for requests
from other machines. The root of the BSD versus System V problem is that
the two listen for different requests. A System V machine can’t send a print
job to a printer attached to a BSD machine, and vice versa.
Task 18.1: Find Local Printers with printers
Of the many problems with printing in UNIX, none is more grievous than trying
to figure out the names of all the different printers available, what kinds of printers
they are, and where they’re located. A complicated configuration file—/etc/printcap—
contains all this information, but it’s definitely not easy to read. So what do you do?
JUST A MINUTE
Some systems have an lpstat command, which lists printers available on
the system. I find the output of this command difficult to read, hence my
inclusion of the printers script here. If you find the output acceptable
(see the next task in this hour for a sample), you can skip this first unit,
although you still might want to spend a few minutes looking at the
printers script anyway.
I will present a simple 20-line shell script, printers, that reads through the /etc/printcap
file and creates an attractive and easily read listing of all printers configured on your system.
This hour presents the script and shows it at work on a few different computer systems. I
encourage you to enter this script and place it in your own bin directory ($HOME/bin should
be in your PATH for this to work).
18
381
1. To start, take a quick look at the contents of the /etc/printcap file:
% head -23 /etc/printcap
# $Header: /usr/msrc/usr/etc/printcap/RCS/printcap,v 1.235 93/11/04
➥10:55:21 mm
Exp Locker: mm $
aglw\ag\Iwag:\
:dr=/usr/local/lib/lp/lpmq:\
:gc=cc:\
:lf=/usr/spool/lpr/aglw/logfile:\
:lo=/usr/spool/lpr/aglw/lock:lp=/dev/null:\
:mj#25:mx#3000:nd=/usr/local/lib/lp/lpnc:\
:pf=gnpt:\
:rm=server.utech.edu:rw:sd=/usr/spool/lpr/aglw:sh:\
:gf=/usr/local/bin/psplot:\
:nf=/usr/local/lib/devps/devps:\
:qo=age:mq=aglw1,aglw2,aglw3,aglw4:mu:\
:wi=AG 23:wk=multiple Apple LaserWriter IINT:
aglw1:\
:dr=/usr/local/lib/lp/lwp.sh:\
:gc=cc:\
:lf=/usr/spool/lpr/aglw1/logfile:\
:lo=/usr/spool/lpr/aglw1/lock:lp=/dev/null:\
:mj#25:mx#3000:nd=/usr/local/lib/lp/lpnc:\
:pf=gnpt:\
:rm=server.utech.edu:rw:sd=/usr/spool/lpr/aglw1:sh:\
:gf=/usr/local/bin/psplot:\
:nf=/usr/local/lib/devps/devps:\
I won’t go into exhaustive detail about the meaning of each field in this listing. It
suffices to say that the first line in each entry lists the name of the printer, a |
character, and any other possible names for the printer. Each field following the
printer name is surrounded by colons and has a two-letter field name (for example,
dr, nf), followed by the value of that particular field or setting. The fields of
interest are the printer name; the wi field, which indicates the location of the
printer; and the wk field, which indicates the type of printer.
2. There are no UNIX utilities to keep you from having to slog through this configuration file. I have written a short, yet powerful, C shell script called printers to list
the desired information in a readable format. Notice the use of a here document to
create the awk script and the multiple-line pipeline at the end of the script that does
all the actual work.
%
#
#
#
#
#
cat bin/printers
printers - create a simple list of printers from the /etc/printcap
file on the system.
From
18
Hour 18
382
set printcap=/etc/printcap
set awkscript=/tmp/awkscript.$$
/bin/rm -f $awkscript
cat << ‘EOF’ > $awkscript
NF == 2 { split($1, words, “|”);
prname=words[1]
}
NF > 2 { printf(“%-10s %s\n”, prname, $0) }
‘EOF’
egrep ‘(^[a-zA-Z]|:wi)’ $printcap | \
sed ‘s/:/ /g’ | \
awk -f $awkscript | \
sed ‘s/wi=//;s/wk=/(/;s/ $/)/’ | \
more
/bin/rm -f $awkscript
exit 0
Some of this script is beyond what you have learned in this book about commands
and scripts. In particular, the awk script, although only four lines long, shows some
of the more powerful features of the program. Enter this as shown, and be careful
to match the quotes and slash characters.
3. Once you’ve entered this script, enter the following:
% chmod +x bin/printers
That will ensure that it’s an executable script. Next, you need to inform the C shell,
using the rehash command, that you have added a new command to the search
path. Then you can try your new shell script:
% rehash
% printers | head -15
aglw
AG 23 (multiple Apple LaserWriter IINT)
aglw1
AG 23 (Apple LaserWriter IINT)
aglw2
aglw3
aglw4
alpslw
LIB 111 (Apple LaserWriter IINTX)
bio
COM B117 (DataPrinter (self-service))
cary
CQuad (NE-B7) (IBM 4019 Laser Printer)
cslw
CS 2249 (Apple LaserWriter IIg)
cs115lw
CS 115 (IBM 4019 LaserPrinter (for CS180))
cs115lw2
CS 115 (IBM 4019 LaserPrinter (for CS180))
csg40lw
CS G040 (IBM 4019 LaserPrinter )
csg50lw
CS G050 (IBM 4019 LaserPrinter )
cslp1
CS G73 (C.Itoh, white paper (self-service))
eng130ci
ENG 130 (C.Itoh, white paper (self-service))
Broken pipe
18
383
You can use this script also to find printers of a certain type or in a specific location, if the descriptions in your /etc/printcap file are configured in the correct
manner:
% printers | grep -i plotter
knoxhp
KNOX 316A (Hewlett Packard 7550+ Plotter)
ccp
MATH G109 (CALCOMP 1073 Plotter)
cvp
MATH G109 (VERSATEC V-80 Plotter)
% printers | grep -i math
lwg186
MATH G186 (Apple LaserWriter IINT(private))
mathci
MATH B9 (C.Itoh, white paper (self-service))
mathlw
MATH 734 (multiple Apple LaserWriter IINT)
mathlw1
MATH 734 (Apple LaserWriter IINT)
mathlw2
mathlw3
cci
MATH G109 (C.Itoh, 3 hole white paper)
ccp
MATH G109 (CALCOMP 1073 Plotter)
cil
MATH G109 (IBM 4019 Laser Printer)
cvp
MATH G109 (VERSATEC V-80 Plotter)
4. You now should be able to choose a printer that’s most convenient for your
location. Set the environment variable PRINTER to that value. You also might want
to tuck that into the last line of your .login file so that next time you log in, the
system will remember your printer selection.
% setenv PRINTER mathlw
% vi .login
If your printer is not responding to what you set the PRINTER variable to,
try using the LPDEST variable, especially on System V.
JUST A MINUTE
setenv NAME “Dave Taylor”
setenv BIN “889”
newmail
mesg y
setenv PRINTER mathlw
~
~
The first, and perhaps biggest, hurdle for printing on UNIX has been solved: figuring
out what the system calls the printer you’re interested in using. Not only do you now
have a new command, printers, for your UNIX system, but you can see how you can
customize UNIX to meet your needs by creating aliases and shell scripts.
18
Hour 18
384
Task 18.2: Printing Files with lpr or lp
Now that you have identified the name of the printer to use, how about sending
information to the printer? If you are on a BSD system, the command to do this is
lpr. You can print the results of a pipe command by adding lpr at the end of the pipeline,
or you can print files directly by specifying them to the program. You can even use < to redirect
input.
If you’re using a System V version of UNIX, you will need to use the lp command instead.
As you read through this hour, you will see the differences between lpr and lp indicated. Note
how the philosophies of the two vary.
The flags available for lpr and lp are numerous, and the most valuable ones are listed in Tables
18.1 and Table 18.2.
Table 18.1. Useful flags for lpr .
Flag
Meaning
-h
Do not print the header page.
Indent the entire file eight spaces before printing.
Print in landscape (sideways) mode, if the printer is capable of doing so.
Send the print job to printer pr.
Print pages in reverse order.
-i
-L
-Ppr
-R
Table 18.2. Useful flags for lp.
Flag
Meaning
-dptr
Send the print job to the printer named ptr.
Print only page n.
Use title as the cover page title, where title is any string.
-Pn
-ttitle
1. Here’s a demonstration of what happens if you try to use lp or lpr without
specifying a printer and without having the PRINTER environment variable set. First,
use the unsetenv command to remove environment variable definitions:
% unsetenv PRINTER
% who | lpr
lpr: No printer specified
Broken pipe
18
385
Some systems default to a printer named lp in this situation, so if you don’t get an
error message, that’s what happened. If you have lpstat (a command for checking
the status of a printer), the -d flag will result in lpstat listing your default printer.
To specify a printer, use the -P flag with lpr or the -d flag with lp, followed
immediately by the name of the printer:
% who | lpr -Pmathlw
Specifying a printer with the -P flag (or -d with lp) always will override the
environment variable specified in PRINTER; therefore, you can specify the default
printer with PRINTER and specify other printers as needed without any further work.
Notice that I printed the output of the who command but received absolutely no
information from the lpr command regarding what printer it was sent to, the print
job number, or any other information.
To make life easier, I’m going to redefine PRINTER:
% setenv PRINTER mathlw
2. To find out what’s in the print queue, I can use lpstat
the lpq -Pprinter command:
-pprinter
on System V or
% lpq -Pmathlw
[email protected]:
driver not active
Printing is disabled.
Pos
--1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
User
---KOSHIHWE
KOSHIHWE
KOSHIHWE
kleimanj
zeta
jharger
jharger
ssinfo
fuelling
zeta
tkjared
SUJATHA
SUJATHA
SUJATHA
bee
bee
bee
ssinfo
ssinfo
ssinfo
ssinfo
stacysm2
Bin
--0104
0104
0104
0317
0042
0167
0167
0353
0216
0042
0142
0043
0043
0043
0785
0785
0785
0353
0353
0353
0353
0321
Size
---008
008
008
032
008
008
008
000
024
152
012
016
024
044
012
056
028
004
000
000
004
000
Jobname
------KOSHIHWE0104a
KOSHIHWE0104b
KOSHIHWE0104c
kleimanj0317a
zeta0042a
jharger0167a
jharger0167b
ssinfo0353a
fuelling0216a
zeta0042b
tkjared0142a
SUJATHA0043a
SUJATHA0043b
SUJATHA0043c
bee0785a
bee0785b
bee0785c
ssinfo0353b
ssinfo0353c
ssinfo0353d
ssinfo0353e
stacysm20321a
18
Hour 18
386
23
24
ssinfo
taylor
0353
0889
000
000
ssinfo0353f
taylor0889a
mathlw: waiting to be transmitted to server.utech.edu
The queue is empty.
Quite a few print jobs are waiting to be sent, but it’s not obvious why the printer is
disabled. The output of the lpq and lpstat commands are explained in detail later
in this hour.
3. To print the file dickens.note in landscape mode, without a header page, indented
eight spaces, and in reverse order, I can use the following flags:
% lpr -hiLR < dickens.note
If I did this often, a C shell alias could be helpful:
% alias lpr...’lpr -hiLR’
On a System V machine, you also could create the alias alias lpr ‘lp’, though
none of these particular options are available with lp.
If you find yourself printing to a couple different printers quite often, you easily
can define a few shell aliases to create printer-specific print commands:
% alias mathprint
% alias libprint
% alias edprint
‘lpr -Pmathlw’
‘lpr -Plibrary’
‘lpr -Pedlw’
On System V machines, the name would be this:
% alias mathprint
% alias libprint
% alias edprint
‘lp -dmathlw’
‘lp -dlibrary’
‘lp -dedlw’
4. Some systems have a command lpinfo that also offers information about printers:
% lpinfo mathlw
mathlw: server.utech.edu; MATH 734; multiple Apple LaserWriter IINT
To find out more information about the printer, you can specify the -v flag:
% lpinfo -v mathlw
mathlw description:
driver: /usr/local/lib/lp/lpmq
printer control group: cc
graphic filter: /usr/local/bin/psplot
log file: /usr/spool/lpr/mathlw/logfile
lock file: /usr/spool/lpr/mathlw/lock
hardware line: /dev/null
maximum job count per user = 25
subqueue list: mathlw1,mathlw2,mathlw3
maximum print file blocks = 3000
make unique via bin change
18
387
network driver: /usr/local/lib/lp/lpnc
ditroff filter: /usr/local/lib/devps/devps
print formats: graphics, ditroff, use pr, troff
queue ordering: age
host attachment: server.utech.edu
spooling directory: /usr/spool/lpr/mathlw
location: MATH 734
description: multiple Apple LaserWriter IINT
5. The lpinfo command also can show you a list of what printers are available, but I
find the output format considerably more difficult to understand than lpstat:
% lpinfo -a | head -15
aglw:
server.utech.edu; AG 23; multiple Apple LaserWriter IINT
aglw1:
server.utech.edu; AG 23; Apple LaserWriter IINT
aglw2:
aglw3:
aglw4:
alpslw: sentinel.utech.edu; LIB 111; Apple LaserWriter IINTX
bio:
ace.utech.edu; COM B117; DataPrinter (self-service)
cary:
franklin.utech.edu; CQuad (NE-B7); IBM 4019 Laser Printer
cslw: server.utech.edu; CS 2249; Apple LaserWriter IIg
cs115lw:
expert.utech.edu; CS 115; IBM 4019 LaserPrinter (for CS180)
cs115lw2:
expert.utech.edu; CS 115; IBM 4019 LaserPrinter (for CS180)
csg40lw:
franklin.utech.edu; CS G040; IBM 4019 LaserPrinter
csg50lw:
franklin.utech.edu; CS G050; IBM 4019 LaserPrinter
cslp1: expert.utech.edu; CS G73; C.Itoh, white paper (self-service)
eng130ci:
age.utech.edu; ENG 130; C.Itoh, white paper (self-service)
Broken pipe
If you find this output readable, you’re undoubtedly becoming a real UNIX expert!
The output of the printers command specifies the location of the printer that printed
the file. I need to go to another building to pick up my hard copy. (The location is
specified in the output of the printers command.)
Task 18.3: Formatting Print Jobs with pr
The printout I generated looked good, but boring. I would like to have a running
header on each page that specifies the name of the file and the page number. I’d also
like to have a bit more control over some other formatting characteristics. This is exactly
where the pr command comes in handy. Not intended just for printing, pr is a general
pagination and formatting command that can be used to display information on the screen.
Even better, pr is available on both BSD and System V UNIX.
The pr program is loaded with options, most of which are quite useful at times. For example,
-2 makes the output two columns, which is useful for printing results of the who command
in landscape mode! The most useful options are presented in Table 18.3.
18
Hour 18
388
Table 18.3. Useful flags in pr.
Flag
Meaning
-n
Produce n-column output per page.
Begin printing on the nth page.
Don’t print the page header and footer information.
Use hdr as the head of each page.
Set the page width to n characters (for landscape mode).
Print all files at once, one per column.
+n
-f
-hhdr
-wn
-m
JUST A MINUTE
On some UNIX systems, the -f flag to pr causes the program to put form
feeds at the bottom of each printed page. To suppress the header and
footer, use -t.
1. My printout of the who command showed me that my choice of paper was poor. In
a 128-character-wide landscape printout, I actually was using only the first 30
characters or so of each line. Instead, I can use pr to print in two-column mode:
% who | pr -2 | more
Dec
9 13:48 1993
root
princess
tempus
enatsuex
coxt
scfarley
nancy
rick
fitzte
maluong
af5
zjin
herbert1
ebranson
billiam
linet2
--More--
18
console
ttyaV
ttyaW
ttyaY
ttyaZ
ttyAa
ttyAb
ttyAc
ttyAd
ttyAe
ttyAg
ttyAh
ttyAi
ttyAj
ttyAk
ttyAm
Page 1
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
6
9
9
9
9
9
9
9
9
9
9
9
9
9
9
9
18:02
13:44
13:43
13:41
13:35
13:36
13:12
13:12
13:47
13:46
09:12
13:44
13:29
13:44
13:36
11:04
ab
dutch
malman
bakasmg
dodsondt
md
rothenba
xuxiufan
nashrm
dls
myounce
liyan
daffnelr
mm
jlapham
chuicc
ttypk
ttypl
ttypm
ttypq
ttyps
ttypv
ttypw
ttypy
ttyq3
ttyq5
ttyq8
ttyq9
ttyqA
ttyqB
ttyqC
ttyqE
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
9
8
9
9
8
8
9
9
9
9
9
9
9
9
9
9
07:57
13:36
13:07
13:09
11:37
08:23
13:15
13:16
13:04
13:06
02:14
13:11
13:36
10:32
12:46
13:38
(nova)
(dov)
(dov)
(age)
(age)
(kraft)
(trinetra)
(ector)
(pc115)
(dialup01)
(limbo)
(volt)
(localhost)
(mm)
(mac18)
(icarus)
389
Notice that the pr program not only made this a two-column listing, but it also
added a page header that indicates the current date and page number.
2. The header still doesn’t contain any information about the command name, which
is what would really be helpful. Fortunately, I easily can add the header information I want by using pr:
% who | pr -h “(output of the who command)” -2 | more
Dec
9 13:50 1993
root
princess
tempus
enatsuex
coxt
scfarley
nancy
rick
fitzte
maluong
maritanj
af5
zjin
herbert1
ebranson
--More--
console
ttyaV
ttyaW
ttyaY
ttyaZ
ttyAa
ttyAb
ttyAc
ttyAd
ttyAe
ttyAf
ttyAg
ttyAh
ttyAi
ttyAj
(output of the who command) Page 1
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
6
9
9
9
9
9
9
9
9
9
9
9
9
9
9
18:02
13:44
13:43
13:41
13:35
13:36
13:12
13:12
13:47
13:46
13:49
09:12
13:48
13:29
13:44
ab
dutch
malman
bakasmg
dodsondt
md
rothenba
xuxiufan
dls
myounce
liyan
daffnelr
mm
jlapham
chuicc
ttypk
ttypl
ttypm
ttypq
ttyps
ttypv
ttypw
ttypy
ttyq5
ttyq8
ttyq9
ttyqA
ttyqB
ttyqC
ttyqE
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
Dec
9
8
9
9
8
8
9
9
9
9
9
9
9
9
9
07:57
13:36
13:07
13:09
11:37
08:23
13:15
13:16
13:06
02:14
13:11
13:36
10:32
12:46
13:38
(nova)
(dov)
(dov)
(age)
(age)
(kraft)
(trinetra)
(ector)
(dialup01)
(limbo)
(volt)
(localhost)
(mm)
(mac18)
(icarus)
That’s much better.
3. I might want to compare the contents of two different directories. Remember that
the -1 flag to ls forces the ls program to list the output one filename per line, so I
can create a couple of files in this format easily:
% ls -1 src > src.listing
% ls -1 /tmp > tmp.listing
These files look like this:
% head src.listing tmp.listing
==> src.listing <==
calc-help
calc.c
fixit.c
info.c
info.o
==> tmp.listing <==
Erik/
GIri/
Garry/
MmIsAlive
18
Hour 18
390
Re01759
Re13201
Sting/
VR001187
VR002540
VR002678
Now I will use pr to build a two-column output:
% pr -m src.listing tmp.listing | head -15
Dec
9 13:53 1993
calc-help
calc.c
fixit.c
info.c
info.o
massage.c
Page 1
Erik/
GIri/
Garry/
MmIsAlive
Re01759
Re13201
Sting/
VR001187
VR002540
Broken pipe
4. This would be more helpful if I could turn off the blank lines automatically
included at the top of each listing page, which is a job for the -f flag (or -t, if your
version of pr was -f for form feeds):
% ^pr^pr -f
pr -f -m src.listing tmp.listing | head -15
Dec 9 13:56 1993
Page 1
calc-help
calc.c
fixit.c
info.c
info.o
massage.c
Erik/
GIri/
Garry/
MmIsAlive
Re01759
Re13201
Sting/
VR001187
VR002540
VR002678
VR002982
VR004477
Broken pipe
5. It looks good. Now it’s time to print by piping the output of the pr command to
the lpr command:
% !pr | lpr
pr -f -m src.listing tmp.listing | head -15 | lpr
18
391
The pr command can be used to ensure that your printouts are always clean and
readable. Again, it’s a perfect place to create an alias: alias print ‘pr | lpr’ or alias
print ‘pr | lp’. Even without any flags, pr automatically adds page numbers to the top of
each page.
Task 18.4: Working with the Print Queue
On a personal computer, you might be used to having your printer directly
connected to your system, so anything you print using PRT: (on DOS) or File | Print
(on the Mac) instantly prints. Unfortunately, UNIX doesn’t grant you the luxury of using
your own personal printer. Instead, it handles print requests in a print queue, a managed list
of files to print. When you send a file to a printer with lpr or lp, the request is added to a queue
of files waiting to print. Your request goes to the bottom of the list, and any subsequent print
requests are added below yours. Your print request gradually moves up to the top of the list
and prints, without interrupting the print requests of those folks ahead of you.
Sometimes it can be frustrating to wait for a printout. However, there are some advantages
to using a queuing system over simply allowing users to share a single printer. The greatest
is that you can use the lprm command to change your mind and remove print requests from
the queue before they waste paper. The lprm command works with the print job name, which
you can learn by checking the print queue, using lpq. Both lprm and lpq either can use the
default PRINTER setting or can have printers specified with -Pprinter. The lpq command also
can limit output to just your jobs by adding your account name to the command.
If your system doesn’t have lprm, use the cancel command to remove entries from the print
queue. The lpstat command is also the System V replacement for the lpq command, though
many sites alias lpq = lpstat to make life a bit easier.
To use cancel, you need to specify the name of the printer and the job ID, as shown in the
lpstat output. If I had print request ID 37 on printer hardcopy, I could cancel the print
request with the command cancel hardcopy -37.
1. A glance at the mathlw queue shows that there are a lot of files waiting to print:
% lpq
[email protected]:
driver not active
Pos
--1
2
3
User
---KOSHIHWE
KOSHIHWE
KOSHIHWE
Bin
--0104
0104
0104
Size
---008
008
008
Jobname
------KOSHIHWE0104a
KOSHIHWE0104b
KOSHIHWE0104c
18
Hour 18
392
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
kleimanj
zeta
jharger
jharger
ssinfo
fuelling
zeta
tkjared
SUJATHA
SUJATHA
SUJATHA
bee
bee
bee
info
info
info
info
stacysm2
info
taylor
0317
0042
0167
0167
0353
0216
0042
0142
0043
0043
0043
0785
0785
0785
0353
0353
0353
0353
0321
0353
0889
032
008
008
008
000
024
152
012
016
024
044
012
056
028
004
000
000
004
000
000
000
kleimanj0317a
zeta0042a
jharger0167a
jharger0167b
ssinfo0353a
fuelling0216a
zeta0042b
tkjared0142a
SUJATHA0043a
SUJATHA0043b
SUJATHA0043c
bee0785a
bee0785b
bee0785c
info0353b
info0353c
info0353d
info0353e
stacysm20321a
info0353f
taylor0889a
The queue is empty.
My print job is job number 24, with the print job name taylor0889a. Figure 18.1
explains the different fields in the queue listing.
Figure 18.1.
The lpq output format
explained.
The printer is also turned off. You can see at the top of the lpq output that telltale
message driver not active Printing is disabled. Obviously, if the printer is
disabled, it’s rather futile to wait for a printout.
2. To limit the output to just those print jobs that are mine, I specify my account
name:
% lpq taylor
[email protected]:
driver not active
18
Pos
--1
User
---taylor
Bin
--0889
Size
---004
393
Jobname
------taylor0889a
The queue is empty.
3. To check the status of another printer, I can specify the printer with the -P flag:
% lpq -Pb280il
[email protected]:
driver not active
The queue is empty.
b280il:
waiting to be transmitted to franklin.utech.edu
The queue is empty.
That’s better. The queue is empty.
4. To remove my print job from the mathlw print queue, I simply specify the print job
name from the lpq output:
% lprm taylor0889a
UNIX carries out my command without giving me confirmation that it has done
so, but a quick check with lpq shows me what’s up:
% lpq taylor
[email protected]:
driver not active
The queue is empty.
The queue is empty.
JUST A MINUTE
I wish the default for the lpq command would show only print jobs that I
have in the queue, and I could use the -a flag to show all print jobs
queued. Furthermore, instead of incorrectly saying The queue is empty,
lpq should report something more useful, like there are 23 other print jobs
in the queue.
5. Now I resubmit the print job request, this time to the b280il printer:
% !pr -Pb280il
pr -f -m src.listing tmp.listing | head -15 | lpr -Pb280il
Uh oh! I don’t want that head
-15
cutting off the information in the printout.
18
Hour 18
394
% lpq -Pb280il
[email protected]:
Pos
--1
2
3
User
---nfsuser
nfsuser
taylor
b280il:
Bin
--0058
0054
0889
Size
---268
012
000
driver active; no job printing
Jobname
------nfsuser0058a
nfsuser0054a
taylor0889a
waiting to be transmitted to franklin.utech.edu
The queue is empty.
To remove my print request, I use lprm:
% lprm taylor0889a
“taylor0889a” not located.
I’ve made a second mistake! I need to specify the printer.
% lprm -Pb280il taylor0889a
Now I can fix the original command and print the files correctly:
% pr -f -m src.listing tmp.listing | lpr -Pb280il
UNIX offers some printing abilities that you might not be accustomed to working
with, particularly the ability to change your mind and stop a print job before it
touches paper. You can see that it’s a good idea to set the PRINTER environment variable to
your favorite printer so that you can save yourself from struggling to enter weird printer
names each time you print a file.
Summary
A few judiciously defined aliases can save you a lot of frustration down the road. Choose your
favorite printer, define the PRINTER environment variable to point to that printer, and give
yourself an alias like print to include all the default options you like for your printouts. You
might consider creating an alias pq to show your own print requests queued for your favorite
printer. (This is easy to do. Use alias pq ‘lpq $LOGNAME’ or alias pq ‘lpstat -u $LOGNAME’.)
You also could show only your print requests, if any, by tucking a grep into the command:
alias pq ‘lpq | grep $LOGNAME’ .
Workshop
in the next hour.
18
395
Key Terms
print job name The unique name assigned to a print job by the lpr or lp command.
print queue The queue, or list, in which all print jobs are placed for processing by the
specific printer.
Questions
1. Use the lpinfo -a or printers command to find out what printers are available on
your system. Which command is easier to use? How many are available?
2. Is your PRINTER variable already set to a printer? Is it the printer you would choose?
3. Use man -k to see what commands you have on your system that work with the
printers and print queues. Use man to peruse them.
4. Show three ways to print the file dickens.note with lpr.
5. Add a print job to the queue and then remove it with lprm. What happened?
6. How would you use pr to add A Tale of Two Cities as a running title across each
printout page of the file dickens.note? How would you start the printout on the
second page of the file?
In the next hour, you learn about the find command, with its unique command flags and
its partner xargs. This command enables you to search the UNIX file system for files that
meet specific criteria, and xargs enables you to perform actions on those files.
18
Hour
397
19
Searching for
Information and Files
One of the greatest challenges in UNIX is to find the files you want, when you
want them. Even the best organization in the world, with mnemonic
subdirectories and carefully named files, can break down and leave you saying
to yourself, “I know it’s somewhere, and I remember that it contains a bid for
Acme Acres Construction to get that contract; but for the life of me, I just can’t
remember where it is!”
Goals for This Hour
■ The find command and its weird options
■ How to use find with xargs
In this hour, you learn sophisticated ways to find specific information on the
UNIX system. The powerful find command and its partner, xargs, are the
contents of this hour.
19
Hour 19
398
Task 19.1: The find Command and Its Weird
Options
The grep family can help you find files by their content. There are a lot of other ways
to look for things in UNIX, and that’s where the find command can help. This
command has a notation that is completely different from all other UNIX commands: It has
full-word options rather than single-letter options. Instead of -n pattern to match filenames,
for example, find uses -name pattern.
The general format for this command is to specify the starting point for a search through the
file system, followed by any actions desired. The list of possible options, or flags, is shown
in Table 19.1.
Table 19.1. Useful options for the find command.
Option
Meaning
-atime n
True if file was accessed n days ago.
True if the file was created n days ago.
Execute command.
True if file was modified n days ago.
True if filename matches pattern.
Print names of files found.
True if file is of type c (as shown in Table 19.2).
True if file is owned by user name.
-ctime n
-exec command
-mtime n
-name pattern
-print
-type c
-user name
The find command checks the specified options, going from left to right, once for each file
or directory encountered. Further, find with any of the time-oriented commands can search
for files more recent than, older than, or exactly the same age as a specified date, with the
specifications -n, +n, and n, respectively. Some examples will make this clear.
1. At its simplest, find can be used to create a list of all files and directories below the
current directory:
% find . -print
.
./OWL
./OWL/owl.h
./OWL/owl
./OWL/owl.c
19
399
./OWL/simple.editor.c
./OWL/ask.c
./OWL/simple.editor.o
./OWL/owl.o
./OWL/Doc
./OWL/Doc/Student.config
./OWL/handout.c
./OWL/owl.question
./OWL/WordMap
./OWL/WordMap/a.out
./OWL/WordMap/lots-of-lines
./OWL/WordMap/msw-to-txt.c
lots and lots of output removed
./src/info.o
./src/massage.c
./keylime.pie
./csh.man
./sample
./sample2
./awkscript
./dickens.note
./newsample
./.sh-history
./mbox
./cheryl
./temp
./temp/zmail
./temp/attach.msg
./.profile
./buckaroo
./sample3
./buckaroo.confused
./deleteme
./dead.letter
./who.is.who
./src.listing
./tmp.listing
./.wrongwords
./papert.article
2. To limit the output to just those files that are C source files (those that have a .c
suffix), I can use the -name option before the -print option:
% find . -name “*.c” -print
./OWL/owl.c
./OWL/ask.c
./OWL/handout.c
./OWL/WordMap/newtest.c
./OWL/feedback.c
./OWL/define.c
./OWL/spell.c
./OWL/submit.c
./OWL/utils.c
19
Hour 19
400
./OWL/parse.c
./OWL/sendmail.c
./owl.c
./src/calc.c
./src/info.c
./src/fixit.c
./src/massage.c
Using the -name option before the -print option can be very handy.
3. To find just those files that have been modified in the last seven days, I can
use -mtime with the argument -7 (include the hyphen):
% find . -mtime -7 -name “*.c” -print
./OWL/owl.c
./OWL/ask.c
./OWL/utils.c
./OWL/sendmail.c
If I use just the number 7 (without a hyphen), I will match only those files that
were modified exactly seven days ago:
% find . -mtime 7 -name “*.c” -print
%
To find those C source files that I haven’t touched for at least 30 days, I use +30:
% find . -mtime +30 -name “*.c” -print
./OWL/WordMap/newtest.c
./src/calc.c
./src/info.c
./src/fixit.c
./src/massage.c
4. With find, I now have a tool for looking across vast portions of the file system for
specific file types, filenames, and so on.
To look across the /bin and /usr directory trees for filenames that contain the
pattern cp, I can use the following command:
% find /bin /usr -name “*cp*” -print
/usr/diag/sysdcp
/usr/spool/news/alt/bbs/pcbuucp
/usr/spool/news/alt/sys/amiga/uucp
/usr/spool/news/comp/mail/uucp
/usr/spool/news/comp/os/cpm
/usr/spool/news/comp/protocols/tcp-ip
/usr/spool/uucp
find: cannot open <“/usr/spool/nqs”>
/usr/spool/lpr/mathcp
/usr/spool/mail/cpotter
/usr/spool/mail/mcpherso
/usr/spool/erpcd/support/acp-config
/usr/spool/erpcd/support/acp-portinfo
/usr/local/bin/cnews/input/recpnews
/usr/local/bin/cppstdin
19
/usr/local/lib/libXdmcp.a
/usr/local/lib/gcc-lib/i386-sequent-bsd4.2/2.4.5/include
➥/netinet/tcp.h
/usr/local/lib/gcc-lib/i386-sequent-bsd4.2/2.4.5/include
➥/netinet/tcp-var.h
/usr/local/lib/gcc-lib/i386-sequent-bsd4.2/2.4.5/cpp
/usr/local/etc/tcpd
/usr/local/etc/acp-restrict
/usr/local/etc/acp-logfile
/usr/local/man/man1/cccp.1
/usr/man/man1/RCS/rcp.1c,v
/usr/man/man1/RCS/cpp.1,v
/usr/man/man1/RCS/cp.1,v
/usr/man/man1/RCS/uucp.1c,v
/usr/man/man1/cpplot.1l
/usr/man/man1/cpio.1u
/usr/man/man1/cp.1
/usr/man/man1/cpp.1
/usr/man/man1/rcp.1c
/usr/man/man1/macptopbm.1u
/usr/man/man1/pbmtomacp.1u
/usr/man/man3/RCS/p-cpus-online.3p,v
/usr/man/man3/RCS/cpus-online.3p,v
/usr/man/man3/RCS/getrpcport.3r,v
/usr/man/man3/cpus-online.3p
/usr/man/man3/getrpcport.3r
/usr/man/man3/p-cpus-online.3p
/usr/man/man3/unitcp.3f
/usr/man/man3/strcpy.3
/usr/man/man3/strncpy.3
/usr/man/man4/RCS/tcp.4p,v
/usr/man/man4/tcp.4p
/usr/man/man8/tcpd.8l
/usr/man/cat3f/%unitcp.3f.Z
/usr/man/cat3f/unitcp.3f.Z
/usr/unsup/bin/cpio
/usr/unsup/gnu/man/man1/cccp.1
/usr/news/cpulimits
/usr/doc/local/form/cp
/usr/doc/local/form/cpio
/usr/doc/local/form/rcp
/usr/doc/uucp
This type of search can take a long time on a busy system. When I ran this
command on my system, it took almost an hour to complete!
JUST A MINUTE
5. To find a list of the directories I’ve created in my home directory, I can use
the -type specifier with one of the values shown in Table 19.2. Here’s one example:
% find . -type d -print
.
./OWL
./OWL/Doc
401
19
Hour 19
402
./OWL/WordMap
./.elm
./Archives
./InfoWorld
./InfoWorld/PIMS
./Mail
./News
./bin
./src
./temp
%
Table 19.2. Helpful find -type file types.
Letter
Meaning
d
Directory
File
Link
f
l
6. To find more information about each of these directories, I can use the -exec
option to find. Unfortunately, I cannot simply enter the command: The exec
option must be used with {}, which will be replaced by the matched filename, and
\; at the end of the command. (If the \ is left out, the C shell will interpret the ; as
the end of the find command.) You also must ensure that there is a space between
the {} and the \:.
% find . -type d -exec ls -ld {} \;
1024 Dec
drwx------ 4 taylor
532 Dec
drwxrwx--- 2 taylor
512 Dec
drwxrwx--- 2 taylor
512 Nov
drwx------ 2 taylor
512 Dec
drwx------ 2 taylor
512 Nov
drwx------ 3 taylor
512 Dec
drwx------ 2 taylor
512 Sep
drwx------ 2 taylor
1024 Dec
drwx------ 2 taylor
512 Oct
drwx------ 2 taylor
512 Dec
drwx------ 2 taylor
512 Oct
drwxrwx--- 2 taylor
512 Nov
10
6
2
7
10
21
3
30
9
6
10
13
8
14:13
18:31
21:18
11:52
13:30
10:39
02:03
10:38
11:42
09:36
13:58
10:45
22:20
.
./OWL
./OWL/Doc
./OWL/WordMap
./.elm
./Archives
./InfoWorld
./InfoWorld/PIMS
./Mail
./News
./bin
./src
./temp
7. The find command is commonly used to remove core files that are more than a
few days old. These core files, as you recall, are copies of the actual memory image
of a running program when the program dies unexpectedly. They can be huge, so
occasionally trimming them is wise:
% find . -name core -ctime +4 -exec /bin/rm -f {} \;
%
19
There’s no output from this command because I didn’t use the -print at the end of
the command.
403
The find command is a powerful command in UNIX. It helps you find files by owner,
type, filename, and other attributes. The most awkward part of the command is the
required elements of the -exec option, and that’s where the xargs command helps
immensely.
Task 19.2: Using find with xargs
You can use find to search for files, and you can use grep to search within files, but
what if you want to search a combination? That’s where xargs is helpful.
1. A few days ago, I was working on a file that was computing character mappings of
files. I’d like to find it again, but I don’t remember either the filename or where the
file is located.
First off, what happens if I use find and have the -exec argument call grep to find
files containing a specific pattern?
% find . -type f -exec grep -i mapping {} \;
typedef struct mappings {
map-entry character-mapping[] = {
int
long-mappings = FALSE;
case ‘l’: long-mappings = TRUE;
if (long-mappings)
/** do a short mapping **/
/** do a long mapping **/
/** Look up the specified character in the mapping database **/
while ((character-mapping[pointer].key < ch) &&
(character-mapping[pointer].key > 0))
if (character-mapping[pointer].key == ch)
return ( (map-entry *) &character-mapping[pointer]);
# map,uucp-map
= The UUCP Mapping Project = [email protected]
grep -i “character*mapping” * */* */*/*
to print PostScript files produced by a mapping application that
➥runs on the
bionet.genome.chromosomes
Mapping and sequencing of
➥eucaryote chromosomes.
./bin/my.new.cmd: Permission denied
typedef struct mappings {
map-entry character-mapping[] = {
int
if (long-mappings)
/** Look up the specified character in the mapping database **/
while ((character-mapping[pointer].key < ch) &&
if (character-mapping[pointer].key == ch)
or lower case values. The table mapping upper to
The output is interesting, but it doesn’t contain any filenames!
19
404
Hour 19
2. A second, smarter strategy would be to use the -l flag to grep so that grep specifies
only the matched filename:
% find . -type f -exec grep -l -i mapping {} \;
./.elm/aliases.text
./Mail/mark
./News/usenet.alt
./src/fixit.c
./temp/attach.msg
3. That’s a step in the right direction, but the problem with this approach is that each
time find matches a file, it invokes grep, which is a very resource-intensive strategy.
Instead, you use the xargs to read the output of find and build calls to grep
(remember that each time a file is seen, the grep program will check through it)
that specify a lot of files at once. This way, grep is called only four or five times
even though it might check through 200 or 300 files. By default, xargs always tacks
the list of filenames to the end of the specified command, so using it is as easy as
can be:
% find . -type f -print | xargs grep -l -i mapping
./.elm/aliases.text
./Mail/mark
./News/usenet.alt
./src/fixit.c
./temp/attach.msg
This gave the same output, but it was a lot faster.
4. What’s nice about this approach to working with find is that because grep is
getting multiple filenames, it will automatically include the filename of any file that
contains a match when grep shows the matching line. Removing the -l flag results
in exactly what I want:
% ^-l^
find . -type f -print | xargs grep -i mapping
./OWL/WordMap/msw-to-txt.c:typedef struct mappings {
./OWL/WordMap/msw-to-txt.c:map-entry character-mapping[] = {
./OWL/WordMap/msw-to-txt.c:int
./OWL/WordMap/msw-to-txt.c:
if (long-mappings)
/** Look up the specified character in
➥the mapping database **/
while ((character-mapping[pointer].key
➥< ch) &&
➥(character-mapping[pointer].key > 0))
19
405
➥if (character-mapping[pointer].key == ch)
return ( (map-entry *)
➥&character-mapping[pointer]);
./.elm/aliases.text:# map,uucp-map
= The UUCP Mapping Project
➥ = [email protected]
./.history:grep -i “character*mapping” * */* */*/*
./.history:find . -type f -exec grep -i mapping {} \;
./Mail/mark:to print PostScript files produced by a mapping
➥application that runs on the
./News/usenet.alt:bionet.genome.chromosomes
Mapping and sequencing
➥of eucaryote chromosomes.
./src/fixit.c:typedef struct mappings {
./src/fixit.c:map-entry character-mapping[] = {
./src/fixit.c:int
./src/fixit.c:
./src/fixit.c:
if (long-mappings)
./src/fixit.c: /** do a short mapping **/
./src/fixit.c: /** do a long mapping **/
./src/fixit.c: /** Look up the specified character in the
➥mapping database **/
./src/fixit.c: while ((character-mapping[pointer].key < ch) &&
./src/fixit.c:
./src/fixit.c: if (character-mapping[pointer].key == ch)
./src/fixit.c:
./temp/attach.msg:or lower case values. The table mapping upper to
When used in combination, find, grep, and xargs are a potent team to help find files
lost or misplaced anywhere in the UNIX file system. I encourage you to experiment
further with these important commands to find ways they can help you work with UNIX.
Summary
The find command is one of the more potent commands in UNIX. It has a lot of esoteric
options, and to get the full power out of find, xargs, and grep, you need to experiment.
Workshop
This Workshop poses some questions about the topics presented in this chapter. It also
provides you with a preview of what you will learn in the next hour.
Questions
1. Use find and wc -l to count how many files you have. Be sure to include the
-type f option so that you don’t include directories in the count.
2. Use the necessary commands to list the following:
■ All filenames that contain abc
■ All files that contain abc
19
406
Hour 19
The next hour introduces you to techniques to communicate with other users.
19
Hour
407
20
Communicating with
Others
It’s time to learn about what’s probably the single most exciting aspect of the
operating system: the ability to communicate with other users on your computer, both interactively and through electronically transmitted mail, e-mail.
Goals for This Hour
■
■
■
■
■
Enabling messages using mesg
Writing to other users with write
Reading electronic mail with mailx
Sending electronic mail with mailx
The smarter alternative for sending mail, elm
Of all the places in UNIX where there is variety, most of it surely is found in
electronic mail, or e-mail. At least 15 different programs are available from
various vendors to accomplish two tasks: to read mail from and send mail to
20
Hour 20
408
other folks. In this hour, you learn about the standard electronic mail system, Berkeley Mail.
I also take a little time to whet your appetite by showing you the Elm Mail System, a fullscreen alternative mail program that’s widely distributed.
JUST A MINUTE
There’s a much bigger world than the machine you’re on; it’s called the
Internet. You learn lots about how to use this valuable system in these
closing hours of Teach Yourself UNIX in 24 Hours.
Task 20.1: Enabling Messages Using mesg
Earlier you learned that all peripherals hooked up to UNIX are controlled by device
drivers and that each device driver has an associated /dev file. If you want to talk with
other users on the system, you need to ensure that they can communicate with you, too. (This
pertains only to write, however; e-mail works regardless of the mesg setting.)
1. To find out through what device I’m connected to the system, I can use the UNIX
command tty:
% tty
/dev/ttyAo
The tty device is just another UNIX file, so I can look at it as I’d look at any other
file:
% ls -l /dev/ttyAo
crw---x--- 1 taylor
21,
71 Dec
8 10:34 /dev/ttyAo*
Notice that I own the file and that I have write permission, but others do not.
2. To enable other users to communicate with me directly, I need to ensure that they
can run programs that can write to my terminal. That is, I need to give them write
permission to my tty device. Instead of using the chmod command—tracking down
what line I’m on and all that—I use a simple alternative, mesg. To turn messages
on—allowing other users to communicate with me—I specify the y flag to mesg:
% mesg y
% ls -l `tty`
crw-rwx--- 1 taylor
21,
71 Dec
8 10:33 /dev/ttyAo*
To disable messages (perhaps if I’m busy and don’t want to be bothered), I can use
the n flag, which says that no, I don’t want messages:
20
% mesg n
% ls -l `tty`
crw---x--- 1 taylor
21,
71 Dec
409
8 10:34 /dev/ttyAo*
3. At any point, you can double-check your current terminal write permission by
entering mesg without any flags. The output is succinct, but it tells you what you
want to know:
% mesg
is n
To see the settings of your tty, use the backquotes with the tty command,
as shown in the preceding examples.
TIME SAVER
Don’t tell anyone this secret. Once you have write permission to someone else’s
terminal, you can redirect the output of commands to their tty device as easily as to
any other file in UNIX. In fact, that’s how the write command works: It opens the other
person’s tty device for writing, and each line you enter is also written to the other person’s
screen. I note this simply so you can see why the permissions of your /dev/tty line are so
important, not so you can go wild and start tormenting your fellow UNIX users!
Task 20.2: Writing to Other Users with write
Now that you can allow others to write to your terminal as well as prevent them from
writing to it, it’s time to find out how to write to theirs and what you can do with
that capability. The command for interacting directly with other users is the write command.
It’s a relatively simple command. When you start write, you specify the other user with whom
you want to communicate, and write starts up for you only, and then it “pages” the other
user to let him or her know that you’re interested in communicating.
Once you’re in the program, each line that you type is sent to the other person as soon as you
press Return. Until they respond by using write on their system to respond, however, they
can’t send any messages to you. Electronic etiquette suggests that you connect and then you
wait without typing until the other user connects with you. Then you can have a
conversation!
To connect with someone, you just need to specify the person’s account name to write. If
the user is logged in more than once, write will try to choose the most recently used line, but
it isn’t always successful. Using w is a good strategy. Simply look at the idle time on each
connection to identify which line the person is actually using. Once you identify the
connection, you can invoke write with the user’s account name and the tty line you desire.
20
Hour 20
410
1. I always start out by ensuring that I’ve turned on mesg to ensure that others can
write to my terminal. Otherwise, the chap at the other end is going to be pretty
darn frustrated trying to talk with me!
% mesg
is y
2. The best way to find out whether your friend is on the system is to use who, piping
the output into the grep program.
% who | grep marv
marv
ttyAx
Dec
8 10:30
He’s logged in on tty line /dev/ttyAx (simply add /dev/ before the line indicated
by who). I can use ls to see whether he has his messages turned on:
% ls -l /dev/ttyAx
crw-rwx--- 1 marv
21,
71 Dec
8 10:33 /dev/ttyAx*
3. To ask him to join a write session, I simply enter the following:
% write marv
What he sees on his screen is the following:
Message from [email protected] on ttyAo at 10:38 ...
_
4. Now I must wait until he responds, which should take only a few seconds. He
types write taylor, and then I see:
Message from [email protected] on ttyAx at 10:40 ...
We’re both connected. Etiquette suggests I wait for his initial hello, which appears
on my screen without preamble:
Hi Dave! -o
_
I can enter lines to him, and he can enter lines to me. When I’m done with my
communication, I press ^d to end it, and he does the same:
Okay, I’ll talk with you tomorrow. -oo
See ya! -oo
^d
EOF
A single press of Return gets me back to the system prompt, %.
20
JUST A MINUTE
411
Because it’s so easy for people to step all over each other’s communication
in write, a simple protocol is borrowed from radio communication: When
you’re done with a transmission (one or more lines of text, in this case),
you should indicate “over,” or -o. Then, the other person types and sends
you information, ending with an -o. When you’re done with the conversation, end with an “over and out,” or -oo. It makes life a lot more pleasant!
Like many things in life, the write command is simple—it has almost no options and
precious little sophistication—yet it is valuable and enjoyable. If you have a quick
question for someone who is logged in, or if you just want to ask your buddy if he’s ready to
have lunch, this is the best way to do it.
Task 20.3: Reading Electronic Mail with mailx
The write command is helpful for those situations when your friend or colleague is
logged in to the computer at the same time you are, but what do you do if the person
is not logged in and you want to leave a note? What if you want a friend to receive a copy of
a note you’re sending to, say, your boss?
That’s where electronic mail moves into the spotlight. Of all the capabilities of UNIX, one
of the most popular is undoubtedly this capability to send electronic mail to another user—
even on another computer system—with a few keystrokes. In this section, you learn how to
work with other users on your own computer, and later in this hour you learn how to send
mail to folks who are on different computers, even in different countries.
A variety of programs for reading mail can be used on UNIX systems, but the two most
common are mail and Mail. (The latter is also often called mailx on SVR4 systems.) Because
of the similarity of the names, the former is known as “mail” and the latter as either “cap mail”
(“cap” for the uppercase M) or “Berkeley Mail.” I refer to “Mail” either as Berkeley Mail or as
its AT&T name, mailx. You should never use mail to read or write mail if Berkeley Mail is
available to you because Berkeley Mail is much easier to use. I will focus on using Berkeley Mail.
To envision electronic mail, imagine that you have a butler who is friendly with the local post
office. You can hand him mail with only the name of the recipient written on the envelope,
and the butler will make sure it’s delivered. If new mail arrives, the butler discreetly lets you
know about it, so you can then display the messages, one by one, and read them. Furthermore,
your butler organizes your old mail in a big filing cabinet, filing each message by any criteria
you request.
That’s almost exactly how Berkeley Mail works. To send mail, you simply state on the
command line the account name of the recipient, indicate a subject, enter the message itself,
and poof! Your message is sent through the system and arrives at the recipient’s terminal
20
Hour 20
412
posthaste. When mail arrives for you, the C shell or one of a variety of utilities, such as biff
or newmail, can notify you. Each time you log in, the system checks for electronic mail, and
if you have any, the system will say You have mail or You have new mail. You can save mail
in files called mail folders.
Berkeley Mail has many command options, both flags that you can specify when you invoke
the program from the command line and commands used within the program. Fortunately,
you always can enter help while you’re in the program to review these options. The most
noteworthy flags are -s subject, which enables you to specify the subject of the message on
the command line, and -f mailfolder, which enables you to specify a mail folder to read
rather than the default (which is your incoming mailbox).
The most valuable commands to use within the program are summarized in Table 20.1.
Table 20.1. Berkeley Mail command summary.
Command
Meaning
delete msgs
Mark the specified messages for deletion.
Display the current page of headers (the cryptic lines of information
at the very top of an e-mail message; I explain them a bit later in
this lesson). Add a + to see the next page, or a - to see the previous
page.
Display a summary of Berkeley Mail commands.
Send mail to the specified address.
Show the specified message or messages.
Leave the Berkeley Mail program.
Respond to the current message.
Save the current message to the specified mail folder.
Undelete the messages that you’ve specified for deletion using the
delete command.
headers
help
mail address
print msgs
quit
reply
save folder
undelete msgs
1. I have lots of electronic mail in my mailbox. When I logged in to the system today,
the shell indicated that I had new mail. To find out what the new messages are, I
use mailx (though I also could have typed Mail because they’re synonymous on my
machine):
% mailx
Mail version 5.2 6/21/85.
20
Type ? for help.
413
“/usr/spool/mail/taylor”: 9 messages 5 new
1 disserli Mon Nov 22 19:40 54/2749 “Re: Are you out there somewhe”
>N 2 Laura.Ramsey Tue Nov 30 16:47 46/1705 “I’ve got an idea...”
N 3 ljw
Fri Dec 3 22:57 130/2712 “Re: Attachments to XALT mail”
N 4 sartin
Sun Dec 5 15:15 15/341 “I need your address”
N 5 rustle
Tue Dec 7 15:43 29/955 “flash cards”
6 harrism Tue Dec 7 16:13 58/2756 “Re: Writing Lab OWL project (“
7 CBUTCHER Tue Dec 7 17:00 19/575 “Computer Based GRE’s”
8 harrism Tue Dec 7 21:46 210/10636 “Various writing environments”
N 9 v892127 Wed Dec 8 07:09 38/1558 “Re: Have you picked up the co”
& _
I have lots of information here. On the very first line, the program identifies itself
as Mail version 5.2, built June 21, 1985. Tucked away in that top corner is the
reminder that I can type ? at any point to get help on the commands.
The second line tells me what mailbox I’m reading. In this case, I’m looking at the
default mailbox for my incoming mail, which is /usr/spool/mail/taylor. On your
system, you might find your mailbox in this directory, or you might find it in a
directory similarly named /usr/mail. Either way, you don’t have to worry about
where it’s located because Berkeley Mail can find it automatically.
The third through eleventh lines list mail messages that I have received from
various people. The format is N in the first column if I haven’t seen the piece of
mail before, a unique index number (the first item in each listing is one), the
account that sent the message, the date and time the message was sent, the number
of lines and characters in the message, and the subject of the message, if known.
Figure 20.1 illustrates this more clearly.
Figure 20.1.
Understanding the
message display in
mailx.
2. To read a specific message, I need enter only the index number of that message:
& 7
Message 7:
From: CBUTCHER Tue Dec 7 17:00:28 1993
From: Cheryl <CBUTCHER>
Subject:
Computer Based GRE’s
To: Dave Taylor <TAYLOR>
I’ve scheduled to take the computer based GRE’s in Indy on Jan. 6th.
Call me crazy but someone’s got to do it. I’ll let you know how it goes.
Do you know anyone else that has taken the GRE’s this way?
there’s a paper in it somewhere.......
I figure
20
Hour 20
414
If you have that handout from seminar in a file, could you please send it
to me?
Thanks.
& _
This message is from my friend Cheryl Butcher. Collectively, the first set of lines in
the message—each a single word, a colon, and some information or other—is the
header of the message, or the electronic equivalent of the postmark and envelope.
The header always will include From:, Subject:, and To:, specifying the name and
electronic address of the sender, the subject of the message, and the list of recipients.
3. To respond to this message, I enter reply:
& reply
To: CBUTCHER
Subject: RE: Computer Based GRE’s
_
Anything I now enter will be sent back to Cheryl:
Hi. I am very interested in hearing about your reaction to the
computer-based GRE test. I’m sure you’re correct that there is
a paper there, but wouldn’t it be best to work with ETS on the
project?
I’ll dig around and find those handouts soonest.
Happy holidays!
Dave
To end the message, I either press ^d on its own line or use the shorthand . by
itself:
.
Cc: _
Berkeley Mail is now asking me to specify any other people I might like to receive
carbon copies of this message. Entering an account name or two here will allow the
designated people to see a copy of this message to Cheryl. Because I don’t want
anyone else to read this message, I press Return, which sends the message and
returns me to the & prompt:
& _
4. I now can use the headers command to see what is the current message (the one I
just read). It’s the message indicated by the >. (Look at Figure 20.1 if you’re having
trouble finding it.)
20
& headers
1 disserli Mon
2 Laura.Ramsey
N 3 ljw
Fri
N 4 sartin
Sun
N 5 rustle
Tue
6 harrism Tue
>
7 CBUTCHER Tue
8 harrism Tue
N 9 v892127 Wed
& _
415
Nov 22 19:40 54/2749 “Re: Are you out there somewhe”
Tue Nov 30 16:47 46/1705 “I’ve got an idea...”
Dec 3 22:57 130/2712 “Re: Attachments to XALT mail”
Dec 5 15:15 15/341 “I need your address”
Dec 7 15:43 29/955 “flash cards”
Dec 7 16:13 58/2756 “Re: Writing Lab OWL project (“
Dec 7 17:00 19/575 “Computer Based GRE’s”
Dec 7 21:46 210/10636 “Various writing environments”
Dec 8 07:09 38/1558 “Re: Have you picked up the co”
To save Cheryl’s message in a folder, I use the save command:
& save cheryl
“cheryl” [New file] 19/575
& _
5. Now that I’m done with this message, I can mark it for deletion with the delete
command:
& delete 7
& _
Notice that after I enter headers, Cheryl’s message vanishes from the list:
& headers
1 disserli Mon
2 Laura.Ramsey
N 3 ljw
Fri
N 4 sartin
Sun
N 5 rustle
Tue
6 harrism Tue
>
8 harrism Tue
N 9 v892127 Wed
& _
Nov 22 19:40 54/2749 “Re: Are you out there somewhe”
Tue Nov 30 16:47 46/1705 “I’ve got an idea...”
Dec 3 22:57 130/2712 “Re: Attachments to XALT mail”
Dec 5 15:15 15/341 “I need your address”
Dec 7 15:43 29/955 “flash cards”
Dec 7 16:13 58/2756 “Re: Writing Lab OWL project (“
Dec 7 21:46 210/10636 “Various writing environments”
Dec 8 07:09 38/1558 “Re: Have you picked up the co”
Look closely at the list, and you will see that it hasn’t completely forgotten the
message; the program hides message 7 from this list. I could still read the message
by using print 7, and I could use undelete 7 to pull it off the deletion list.
JUST A MINUTE
Deleted messages in Berkeley Mail are actually marked for future deletion
and aren’t removed until you quit the program. Once you quit, however,
there’s no going back. A deleted message is gone. While you’re within
the program, you can delete and undelete to your heart’s content.
6. Now I want to delete both messages from harrism (numbers 6 and 8):
& delete 6 8
Now the list of messages in my mailbox is starting to look pretty short:
& h
1 disserli Mon Nov 22 19:40 54/2749 “Re: Are you out there somewhe”
2 Laura.Ramsey Tue Nov 30 16:47 46/1705 “I’ve got an idea...”
20
Hour 20
416
N
N
N
>N
& _
3
4
5
9
ljw
sartin
rustle
v892127
Fri
Sun
Tue
Wed
Dec
Dec
Dec
Dec
3
5
7
8
22:57
15:15
15:43
07:09
130/2712 “Re: Attachments to XALT mail”
15/341 “I need your address”
29/955 “flash cards”
38/1558 “Re: Have you picked up the co”
Most commands in Berkeley Mail can be abbreviated to just their first
letters, which cuts down on typing.
TIME SAVER
7. You can save a group of messages to a file by specifying the numbers between the
save command and the folder name:
& save 6 8 harris
6: Inappropriate message
Oops. I deleted messages 6 and 8. I must undelete them before I can proceed:
& undelete 6 8
& save 6 8 harris
“harris” [New file] 268/13392
8. Use the quit command to get out of this program:
& quit
Saved 1 message in mbox
Held 6 messages in /usr/spool/mail/taylor
%
The messages that I viewed and didn’t delete are moved out of my incoming
mailbox to the file mbox. The messages that I saved and the messages I marked
for deletion are silently removed, and all remaining messages are retained in
/usr/spool/mail/taylor.
JUST A MINUTE
The biggest complaint I have with Berkeley Mail is that it does all this activity
silently. I don’t like the fact that saved messages are deleted automatically
from the incoming mailbox when I quit and that—more importantly—
messages I’ve read are tossed automatically into another folder. To ensure
that messages you’ve read aren’t moved into mbox when you quit, you can
use the preserve command, which you can use with a list of numbers, the
same way you can use other Berkeley Mail commands. Any message with
which you use preserve will remain in your incoming mailbox.
Once you get the hang of it, Berkeley Mail offers quite a lot of power, enabling you
to read through your electronic mail, save it, and respond as needed with ease. The
program has considerably more commands than shown here, so further study is helpful.
20
417
Task 20.4: Sending Mail with mailx
Now you know how to read your electronic mail using Berkeley Mail (aka mailx),
and you know how to send mail from within the program. How do you send
messages and files to people from the command line? It’s quite simple. You even can specify
the message subject with the -s starting flag.
1. To send a message to someone, enter the name of the command, followed by the
recipient’s account name:
% mail marv
Subject: Interested in lunch tomorrow?
_
I now can enter as many lines of information as I want, ending, as within the
Berkeley Mail program itself, with either ^d or .:
I’m going to be in town tomorrow and would like to
rustle up some Chinese food. What’s your schedule
look like?
Dave
.
Cc: _
Again, I’m offered the option of copying someone else, but—again—I opt not to
do so. Pressing Return sends the message.
2. To send a file to someone, combine file redirection with the use of the -s flag:
% mail -s “here’s the contents of sample.file” marv < sample.file
The file was sent without any fuss.
3. Even though Berkeley Mail gives you no indication, several commands are available
for use while entering the text of a message, and all can be listed with ~?:
% mail dunlaplm
Subject: Good morning!
~?
---------------------------------------------------------The following ~ escapes are defined:
~~
Quote a single tilde
~b users
Add users to “blind” cc list
~c users
Add users to cc list
~d
Read in dead.letter
~e
Edit the message buffer
~h
Prompt for to list, subject and cc list
~m messages
Read in messages, right shifted by a tab
~p
Print the message buffer
~r file
Read a file into the message buffer
~s subject
Set subject
20
Hour 20
418
~t users
Add users to to list
~v
Invoke display editor on message
~w file
Write message onto file.
~?
Print this message
~!command
Invoke the shell
~|command
Pipe the message through the command
---------------------------------------------------------_
The ones most important to remember are ~v, to start up vi in the message; ~r, to
read in a file; ~h, to edit the message headers; ~!, to invoke a shell command; and
~p, to show the message that’s been entered so far:
I wanted to wish you a cheery good morning! You asked about
the contents of that one file, so here it is:
~!ls
Archives/
bin/
deleteme
sample
InfoWorld/
buckaroo
dickens.note
sample2
Mail/
buckaroo.confused keylime.pie
src/
News/
cheryl
mbox
temp/
OWL/
csh.man
newsample
awkscript
dead.letter
owl.c
!
The output of the command isn’t included in the message, as is shown if you use
the ~p command:
~p
------Message contains:
To: taylor
Linda,
I wanted to wish you a cheery good morning!
(continue)
_
You asked about
4. To read in a file, use the ~r command:
~r dickens.note
“dickens.note” 28/1123
Here, the contents of the file are included in the note, but mailx didn’t list the
contents to the screen. Again, using ~p will list the current message:
------Message contains:
To: taylor
Linda,
20
I wanted to wish you a cheery good morning!
You asked about
419
Preface
care and interest.
Tavistock House
November 1859
(continue)
5. I can fine-tune the headers using the ~h command:
~h
To: dunlaplm_
Pressing Return leaves it as is, and pressing Backspace lets me change it as desired.
A Return moves to the next header in the list:
Pressing Return a few more times gives me the opportunity to change other headers
in the message:
Cc:
Bcc:
(continue)
The Cc: header allows me to specify other people to receive this message. The Bcc:
is what’s known as a blind carbon copy, an invisible copy of the message. If I send a
message to dunlaplm and a carbon copy to cbutcher, each can see that the other
received a copy because the message will have To: dunlaplm as a header and also will
list the other’s name after Cc:. If I want to send a copy to someone without any of the
other parties knowing about it, that’s where a blind carbon copy can be helpful.
Specifying someone on the Bcc: list means that that person receives a copy of the
message, but his or her name doesn’t show up on any header in the message itself.
20
Hour 20
420
6. Finally, I use ^d to end the message.
^d
Cc:
%
All so-called tilde commands (so named because they all begin with the ~, or tilde,
character) are available when you send mail from the command line. They also are
available when you send mail while within the Berkeley Mail program.
Task 20.5: The Smarter Electronic Mail
Alternative, elm
Just as line editors pale compared to screen editors such as vi, so does Berkeley Mail
when compared to the Elm Mail System, or elm. Although it’s not available on all
UNIX systems, the Elm Mail System is widely distributed, and if you don’t have it on your
system, your system’s vendor should be able to help out.
The basic premise of elm is that the user should be able to focus on the message, not the
medium. Emphasis is placed on showing human information. The best way to show how it
works is to go straight into it!
I’m probably just a bit biased about elm because I am the author of the
program. The widespread acceptance of the design, however, suggests that
I’m not alone in having sought a friendlier alternative to Berkeley Mail.
JUST A MINUTE
Another mailer with a very similar user interface is Pine. If you have
access to both Elm and Pine, however, I recommend that you pick Elm
because it lets you work with your mail in a much more efficient manner.
1. To start up the Elm Mail System, enter elm:
% elm
The screen clears and is replaced with this:
Mailbox is ‘/usr/spool/mail/taylor’ with 15 messages [ELM 2.3 PL11]
--> 1
2
3
4
5
20
Dec
Dec
Dec
Dec
Dec
8
7
7
7
7
v892127@nooteboom.
Mickey Harris
Cheryl
Mickey Harris
Russell Holt
(52)
(214)
(24)
(69)
(37)
Re: Have you picked up the com
Various writing environments
Re: Writing Lab OWL project
flash cards
6
7
8
9
OU 10
Dec
Dec
Dec
Dec
Dec
7
5
5
4
3
Bill McInerney
Mickey Harris
Rob Sartin
J=TAYLOR@MA@168ARG
Linda Wei
(121)
(29)
(31)
(28)
(143)
421
New Additions to U.S. Dept
Re: OWL non-stuff
I need your address
Note to say HI!
Re: Attachments to XALT
You can use any of the following commands by pressing the first char
d)elete or u)ndelete mail, m)ail a message, r)eply or f)orward, q)uit
To read a message, press <return>. j=move down, k=move up, ?=help
Command: _
The current message is indicated by the arrow (or, on some screens, the entire
message line appears in inverse video). Whenever possible, elm shows the name of
the person who sent the message (for example, Mickey Harris rather than mharris
as in Berkeley Mail), indicates the number of lines in the message (in parentheses),
and shows the subject of the message.
The last few lines on the screen indicate the options available at this point. Notice
that j and k move the cursor up and down the list, just as they move up and down
lines in vi.
2. To read a message, use the j key to zip down to the appropriate message and press
Return. You then will see this:
Message 3/15
From Cheryl
Dec 7 ’93 at 4:57 pm est
I’ve scheduled to take the computer based GRE’s in Indy on Jan. 6th.
Call me crazy but someone’s got to do it. I’ll let you know how it
goes.
Do you know anyone else that has taken the GRE’s this way?
there’s a paper in it somewhere.....
I figure
If you have that handout from seminar in a file, could you please
send it to me?
Thanks.
Command (‘i’ to return to index): _
20
Hour 20
422
At this point, you can use j to read the next message directly, r to reply, or i to
return to the table of contents.
3. I realized that I said something in my message to Cheryl that was incorrect. I can
type r here to reply to her message. Typing r causes the last few lines of the screen
to be replaced with this:
-----------------------------------------------------------------------Command: Reply to message
Copy message? (y/n) n
To include the text of the message in your response, type y. I don’t want to, so I
press Return:
To: CBUTCHER (Cheryl)
Subject of message: Re: Computer Based GRE’s_
Now you can see the address to which the response will be sent, the name of the
recipient (in parentheses), and the subject of the message. (The elm command
automatically adds the Re prefix to the subject.) The cursor sits at the end of the
subject line so you can change the subject if you wish. It’s fine, so I again press
Return:
To: CBUTCHER(Cheryl)
Subject of message: Re: Computer Based GRE’s
Copies To: _
No copies are needed, so I again press Return. The bottom of the screen now looks
like this:
To: CBUTCHER(Cheryl)
Subject of message: Re: Computer Based GRE’s
Copies to:
Enter message. Type Elm commands on lines by themselves.
Commands include: ^D or ‘.’ to end, ~p to list, ~? for help.
_
20
423
Notice that ~p and ~? are available. In fact, all the tilde commands available in
Berkeley Mail also are available in the Elm Mail System.
I enter the message, and end with a .:
Just a reminder that we have that seminar tomorrow
afternoon too. See ya there?
-- Dave
.
Ending the message calls up this:
Please choose one of the following options by parenthesized letter: s
e)dit message, edit h)eaders, s)end it, or f)orget it.
I press Return once more, and the message is sent.
4. I type i to return to the index page and q to quit.
There’s a lot more that the Elm Mail System can do to simplify your electronic mail
interaction. If elm is available on your system, I encourage you to check it out further,
and if you don’t, try calling your vendor or a user group to see if someone else can arrange
for you to have a copy. Like the Free Software Foundation applications, elm is free. With it
you even get the source so you can see how things are done internally if you’re so inclined.
Summary
For awhile, you’ve known that there are other users on your computer system, and you’ve
even learned how to find out what they’re doing (with the w command). Now you know how
to communicate with them, too!
Here’s a word of advice: It can be frustrating and annoying to be pestered by unknown folk,
so I recommend that you begin by sending mail to yourself and then to just your friends on
the system. After some practice, you’ll learn how net etiquette works and what is or isn’t
appropriate for write or mail.
TIME SAVER
You can learn a lot more about network etiquette by visiting my online
Network Etiquette Primer at http://www.intuitive.com/tyu24/
netiq.html .
Workshop
in the next hour.
20
Hour 20
424
Key Terms
blind carbon copy
recipient.
An exact copy of a message, sent without the awareness of the original
carbon copy An exact copy of a message sent to other people. Each recipient can see the
names of all other recipients on the distribution list.
e-mail
Electronically transmitted and received mail or messages.
mail folder A file containing one or more e-mail messages.
mail header The To:, From:, Subject:, and other lines at the very beginning of an e-mail
message. All lines up to the first blank line are considered headers.
mailbox A synonym for mail folder.
preserve Ensure that a message doesn’t move out of your incoming mailbox even though
you’ve read it.
starting flag Parameters that you specify on the command line when you invoke the program.
tilde command A command beginning with ~ in Berkeley Mail or the Elm Mail System.
undelete
Restore a deleted message to its original state.
Questions
1. Use tty to identify your terminal device name, and then use ls to look at its current
permissions. Do you have messages enabled or disabled? Confirm with the mesg
command.
2. Try using the write command by writing to yourself; or, if you have a friend on the
system, try using write to say hi and see whether the person knows how to respond.
If he or she doesn’t respond in about 30 seconds, you might want to enter To
respond to me, type write joe at the command line! (filling in your account
name in place of joe).
3. Send yourself a message using mailx.
4. Now use Berkeley Mail to read your new message, and then save it to a file, delete
it, undelete it, and save it to a mail folder.
5. Start Berkeley Mail so that it reads in the newly created mail folder rather than in
your default mailbox. What’s different?
6. If elm is available to you, try using it to read your mail. Do you like this mail
program or Berkeley Mail better? Why?
20
In the next hour, you learn about the World Wide Web and how to use Netscape Navigator
to access Web sites throughout the world.
Hour
425
21
Using Netscape To See
the World Wide Web
In the previous hour, you learned how to communicate across the Internet with
other users via talk and mail. This hour introduces you to the World Wide
Web, the most commonly recognized part of the Internet.
Goals for This Hour
■
■
■
■
The basics of the Internet
How to start your browser
How to find some sites
How to customize your browser
The Internet and its predecessors have been around for as long as, or even longer
than UNIX. Originally, the Defense Advanced Research Projects Agency,
DARPA, decided to build a computer network for national defense. The
National Science Foundation also encouraged and funded a network of computers for academic research. As scientists became aware of UNIX, many of
21
426
Hour 21
these network applications were ported. Because the UNIX operating system was itself
designed to be portable between machine architectures, UNIX quickly became the operating
system of choice for networking, and it became the backbone of what is now called the
Internet.
The most common way to access the Internet today is through a Web browser. The two most
common browsers are Netscape’s Navigator and Microsoft’s Internet Explorer.
Although Microsoft makes a version of its browser for UNIX, Navigator is by far the most
popular UNIX browser.
This hour introduces you to the World Wide Web using Netscape Navigator.
Introduction to the Internet
The Internet is a very interesting organization of computers. When most people hear the
word Internet, they immediately think of the World Wide Web and Web browsers. The
Internet is really a lot more than that.
The first use of the Internet’s predecessors was to transfer data from one machine to another,
primarily for either defense-related applications (such as missile targeting, code breaking, and
the like) or scientific research. Early in the Internet’s life, the capability to send messages
between users was added. This has since developed into what is now called electronic mail,
or e-mail. E-mail messages were exchanged initially back in the late 1960’s.
Two subsequent additions were the capabilities to transfer files and to remotely access
machines. These capabilities, defined as protocols, have evolved into FTP (for transferring
files) and Telnet (for remotely accessing machines). Protocols are the underlying methods on
the Internet that enable two different machines to exchange information. The UNIX tools
that implement these protocols are described in greater detail in the next hour.
In the late 1970s, two graduate students, Tom Truscott at Duke University and Steve
Bellovin at the University of North Carolina, developed a means to transmit general-purpose
messages between the two campuses. This has since evolved into Netnews, which is essentially
a bulletin board system (BBS) where messages are not stored on a single machine but are
shared between all the machines that contact the BBS.
The 1980s saw the addition of the Internet Relay Chat (IRC) protocol. These are text
messages shared instantly with other people subscribed to a discussion. This is where you can
take part in real-time communication with others.
Another protocol was introduced by the University of Minnesota, and in homage to the
university’s mascot, it is naturally called Gopher. This was a means of making files available
to browsers anywhere on the Web. Details on using gopher are in Hour 22, “Internet E-Mail,
Netnews, and IRC.”
21
427
All of these remained primarily text oriented. Sure, there is a standard, called MIME
(multimedia Internet mail extension), for sending pictures, sounds, and other non-text files
via e-mail and Netnews, but its use is still rather restricted. Even so, MIME requires the
translation of the images into a form of text that is again translated at the receiving end. There
was still a need for a method to transmit sound, images, movies, and other data via the
Internet. These problems have been addressed in the World Wide Web.
Underneath the Web is a protocol called HTTP, for hyper-text transfer protocol. This protocol
allows for the transmission of non-text files for images, as well as text for pages. HTTP is the
underpinnings of Web communications.
On the presentation level is a language called HTML, for hypertext markup language. This
is an interpreted mark-up language that specifies layout and presentation of information. The
file that you create will not look like the file displayed. A markup language is one where the
formatting instructions of a document are actually text that is interspersed (and visible)
throughout the document, and you don’t see the effects of them until the document is
interpreted by a printer or browser. In WYSIWYG word processers, such as Word, the
formatting instructions are in the document, but you don’t see them.
Web browsers, however, go far beyond using HTTP. You can specify in HTML different
links (special sections of a document that you can use to access other documents), using a
fairly large number of protocols. Mail, Netnews, FTP, Telnet, Gopher, and HTTP all are
supported by most browsers. This tends to make the browser a universal tool for accessing
many of the resources on the Internet.
Task 21.1: Starting Your Browser
This section introduces you to the basics of Netscape Navigator. Navigator has many
command-line options, but the primary methods you learn here are for surfing the
Net, starting at a specific site, or examining your own local files.
Before you get started with Navigator, you need to be running the X Window
System, which I refer to as “X.” X is the standard graphical interface for most UNIX
systems and usually can be started with the command startx. On my system, you won’t even
need to do that because once you log in, you are already running X. If you can move your
mouse on a UNIX system, odds are you have the X Window System installed.
If you can’t run X, talk with your system administrator or your software provider. You’ll be
amazed how much better UNIX looks when you have a graphical interface to help out!
The next step is to enter the command netscape at the shell prompt. This first either brings
up a license agreement, as shown in Figure 21.1 (if this is the first time you’ve run Navigator),
or just brings up the main browser window. After a short period of time (long enough for you
to read all the legal mumbo-jumbo), this window is replaced by the Netscape home page (see
Figure 21.2).
21
Hour 21
428
Figure 21.1.
The Navigator license
agreement.
Figure 21.2.
The Netscape home page.
JUST A MINUTE
If Navigator doesn’t start for you, it may not be installed. Read about
ftp in the next hour, and then ftp a copy of Navigator from
ftp.netscape.com and install it. Be certain to get the right Navigator
for your hardware.
If you have problems doing this, talk with your system administrator.
21
JUST A MINUTE
429
You may not end up on the Netscape home page. This is entirely dependent on the configuration set up by the software provider. For example, on
BSDI UNIX (a BSD product for Intel Computers distributed by Berkeley
Systems Design, Inc.), the default home page is http://www.bsdi.com ,
their corporate home page. You can change your default starting page
pretty easily from the Preferences section of the program.
When you reach this window, you can click on the text that is underlined in blue (these are
the hypertext links on the Web page), or you can click on buttons displayed on the Web page
or buttons displayed by the browser itself to go to new locations.
A good place to start is the What’s New page (see Figure 21.3). Click on the What’s New?
button in your browser to get a listing of new sites.
Figure 21.3.
Netscape’s What’s New
site.
Being interested in wildlife, I’ll click on the “Arabian Wildlife Online” link to visit that site,
as shown in Figure 21.4.
You can continue from there to explore other sites.
Another means of starting Navigator is by specifying a URL on the command line. This starts
Navigator at the specified location:
% netscape http://www.internetmall.com
21
Hour 21
430
Figure 21.4.
The Arabian Wildlife
Online Web site.
This particular command sends you to the largest location for shopping on the Internet, as
shown in Figure 21.5, and you can explore the different stores from there. If you want to see
how people are making money on the Net, this is a good place to start!
Figure 21.5.
Shop ’til you drop at The
Internet Mall.
21
The last method to start Navigator is by specifying a local file on the command line. If the
file is HTML, it will be loaded, and you will see it displayed in the browser. (This is a good
method for testing pages you are writing without making them publicly available.) If the file
431
is text, it is displayed purely as text, and if it’s a graphic of either type XBM (an X bitmap
format), GIF (Graphics Interchange Format, probably the most common form of graphics
on the Web), or JPEG (Joint Photographic Experts Group, a high-color format), the graphic
will be displayed against a gray background in the browser.
I’ve been working on a site called Raptor World and recently wrote the page describing the
bald eagle (see Figure 21.6). I could check and see how it looks simply by typing netscape
followed by the name of the file, baea.html:
% netscape baea.html
Figure 21.6.
You also can view local
files with Netscape.
You may note the strange graphic under the title, “Bald Eagle.” This is a placeholder for a
picture that I plan to add later. It isn’t there yet, so the browser doesn’t know what to place
there. This and other changes need to be made to the page before I bring up the page online.
If you find a site you are interested in visiting, you can pull down the Bookmark menu and
add the site to your list of bookmarks. Bookmarks are sites that you have decided are
interesting and that you want to save for easy recall. Then, you can later pull down that menu
and go straight to that site.
There are three simple methods for getting started with Navigator. You can first visit
your default home page and surf the Web from there, you can go to a specific location
as your starting point, or you can use Navigator to examine files on your home system.
21
Hour 21
432
Task 21.2: Finding Some Sites
One of the biggest weaknesses in the Web is the lack of organization. People who
write Web pages try to keep their own sites organized, but the volume of information
out there is large, and finding the information you want can be difficult.
Here, I set out with a plan. I’m a fan of Duke basketball, so I want to find out as much
information about the team as there is available. There are several sites that have
search engines to find information on the Web. One of the oldest is Yahoo!, http://
www.yahoo.com/, so let’s start there. Yahoo! is shown in Figure 21.7.
Figure 21.7.
Despite the weird name,
Yahoo! is very serious
about searching the Web.
The most likely search location is Sports, under Recreation and Sports. So, I’ll click on that
and see what information is there (see Figure 21.8).
As you can see, basketball is listed—with over 2,600 links! If I click on that link, I see the
screen displayed in Figure 21.9.
21
433
Figure 21.8.
Use Yahoo!’s tree
structure to drill down to
the topic you want.
Figure 21.9.
Each topic breaks down
into more diverse subtopics.
The next stop is obviously College and University, which takes me to Figure 21.10.
21
Hour 21
434
Figure 21.10.
Finally, the goal is close
at hand.
And from there to Men, then Teams, which reveals what’s in Figure 21.11.
Figure 21.11.
After a long road, the
goal is reached.
Finally! I see Duke Blue Devils listed. If I click this, I find there are seven links (see Figure
21.12).
21
435
Figure 21.12.
Many links are crossreferenced under different
topics.
As shown in Figure 21.12, the title has changed to refer to all Duke athletics. Most of the pages
refer to basketball. The most interesting page looks like Duke Basketball Report (see Figure
21.13), so I’ll click there.
Figure 21.13.
Welcome to North
Carolina.
21
This page is good, so I’ll set a bookmark to it. To do this, pull down the Bookmarks menu
from the title bar and select Add Bookmark.
Hour 21
436
Another search method is to use a site that has a Web crawler to index pages. Web crawlers
are specialized programs that search out Web sites and attempt to index them.
AltaVista is one such Web crawler. So, I first go to their home page, http://www.altavista.
digital.com/ (see Figure 21.14).
Figure 21.14.
The AltaVista home
page.
This is a search system, so I enter Duke basketball in the Submit box. This produces a much
longer list of sites (see Figure 21.15), but on the first page, I see 10 sites, including the Duke
Basketball Report I just bookmarked.
Figure 21.15.
The search results.
21
437
I can search over this list and find other places that aren’t registered with Yahoo!. I am
particularly interested in the page with “The Shot.” So, I click on that page and find a set of
pictures from Duke’s recent history. One is of the exultation after Christian Laettner hit the
shot to beat Kentucky in overtime in 1992, and one is of the back-to-back NCAA
championship banners. A great site.
Finding information on the Web can be tricky, but many sites exist that attempt to
catalog Web sites. Of the several starting points, Yahoo! and AltaVista provide
copious volumes of data.
Task 21.3: Customizing Your Browser
Once you’ve been using Navigator for a while, you may want to change how it
appears. Fortunately, Navigator is quite customizable, by using the Options menu.
After you have started Navigator, click on the Options button on the menu bar. At the top
of the menu are five areas for customization. The one I’ll look at today is General Preferences.
There are five tabs under General Preferences that provide areas for customization:
Appearance, Fonts, Applications, Helpers, and Images, as shown in Figure 21.16. Under
Appearance, you can change your default home page. Because I liked that Duke Basketball
Report page, I can enter it. By clicking OK, I make that my default home page.
Figure 21.16.
Navigator’s General
Preferences window.
Under Fonts (see Figure 21.17), I find a list of different fonts available and in different sizes.
Here, I’ve changed the font size to be a bit larger and to use New Century Schoolbook as my
typeface. When I look at my home page with this new font setting, it looks like Figure 21.18.
21
Hour 21
438
Figure 21.17.
Changing your font
settings.
Figure 21.18.
Customize the look of
every Web page you visit.
The next option is for Applications (see Figure 21.19). These are the commands started by
Navigator when a protocol is requested. I’ll leave these intact. If you have a different UNIX
configuration than the default, you may want to change them. Check with your administrator.
21
439
Figure 21.19.
You can use any application to view sites out on
the Net.
Helpers (shown in Figure 21.20) are tools to interpret MIME input, the mechanism by which
Web servers send different types of Web page elements such as graphics, audio, and so on.
I would recommend extreme caution before you modify any of these settings lest you end up
being unable to display common Web page elements.
Figure 21.20.
The list of helper
applications.
The last tab under General Preferences is Images, shown in Figure 21.21. These define how
images are displayed; because I’d rather not see an incomplete image, I click Display Images
After Loading.
21
Hour 21
440
Figure 21.21.
Set up how images are
displayed.
Navigator is fully customizable from the Options menu on your browser. Although
normally it is set up for the standard UNIX configuration, if you have differences
from the norm, you will need to customize Navigator.
Summary
This hour introduced you to the basics of Netscape Navigator and the World Wide Web.
Now you know how to start and use Navigator to access resources on the Internet and how
to find resources through different search engines.
Workshop
in the next hour.
Key Terms
bookmarks A listing of favorite sites for quick retrieval.
browser A program designed to load hypertext pages and follow hyperlinks.
hyperlinks Specifications within a document that include instructions for loading a
different document.
surfing A style of interacting with the World Wide Web, usually for pleasure, where you
follow hyperlinks from Web site to Web site.
21
441
URL The specification for a document on the World Wide Web. Usually, it includes a
protocol, machine name, and filename.
World Wide Web A collection of sites that provide hypertext documents on the Internet.
Questions
1. How would you find sites about General Custer?
2. Where is Sagarmatha National Park?
3. How would you change your default home page?
In the next hour, we look at different Internet tools, including Internet e-mail and Netnews.
21
443
22
Hour
22
Internet E-Mail,
Netnews, and IRC
I thought you might enjoy a guided tour of some of the more astounding
resources on the Internet, a global network of UNIX and other computer
systems—all connected by high-speed links. Most of these resources require you
to have an account on a machine that’s connected to the Internet, but a
surprising number of them have electronic mail alternatives, too, so you won’t
be completely in the cold even if you don’t have an account. If you don’t have
access to the Internet, I encourage you to read this final hour just to get a feel
for the astounding amount and variety of information and services already
available on the fastest growing network in the world.
Goals for This Hour
■ Sending e-mail to Internet users
■ Talking with remote Internet users
■ Searching databases with WAIS
Hour 22
444
■ Having the whole world with gopher
■ All the news that’s fit or otherwise
This hour is intended to offer a quick and enjoyable overview of the many services available
through the Internet. From finding that long-lost archive to looking for a book at a library
overseas, the range of information available undoubtedly will astound you!
Task 22.1: Sending E-Mail to Internet Users
The most common use of the Internet is probably to send electronic mail between
individuals and to mailing lists. What’s really a boon is that everyone, from New
York to Los Angeles, Japan to Germany, South Africa to India, has an address that’s very
similar, and you’ve already seen it shown here! The notation is [email protected], where user
is the account name or full name, host is the name of the user’s machine, and domain is the
user’s location in the world.
By reading the host and domain information from right to left (from the outside in, really),
you can decode information about someone by looking at the person’s e-mail address. My
address at a system called Netcom, for example, is [email protected], which,
reading right to left, tells you that I’m at a commercial site (com), with a company by the name
of Netcom (netcom), and the name of the computer I’m using is netcom2. My account on
Netcom is taylor.
There are lots of top-level domains, and the most common are shown in Table 22.1.
Table 22.1. Common top-level Internet domains.
Domain
Type of Site or Network
edu
Educational sites
Commercial businesses
Military or defense systems
Alternative networks accessible via Internet
Nonprofit organizations
United States systems not otherwise classified
com
mil
net
org
us
1. To send mail to someone on the Internet is easy because the C shell doesn’t view
the @ as a special character. If you’d like to send me a message, for example, you
could use this:
% mailx [email protected]
Subject: _
22
Enter the message and end it with a ^d as you would in any e-mail message. It is
immediately sent to me!
JUST A MINUTE
I encourage you to drop me a note if you’re so inclined, letting me know
how you’re enjoying this book, any problems you might have encountered, and any commands you were puzzled by that might be easier with
a bit more explanation. If nothing else, just say hi!
2. Although electronic mail addresses always follow the same format, they can vary
quite a bit. To give you an idea of the variation, I used grep to extract the From:
addresses of some mail I’ve recently received:
% grep ‘^From:’ /usr/spool/mail/taylor
From: Steve Frampton <[email protected]>
From: Joanna Tsang <[email protected]>
From: “Debra Isserlis” <[email protected]>
From: “Jay Munro [PC Mag]” <[email protected]>
From: [email protected] (Linda Wei)
From: Cheryl <[email protected]>
From: [email protected] (Mickey Harris)
From: [email protected]
From: “ean houts” <[email protected]>
From: [email protected] (Mickey Harris)
From: “Barbara Maxwell” <[email protected]>
From: [email protected] (Steve Mansour)
From: [email protected] (Aditya Bhasin)
From: [email protected] (Marvin Raab)
The national convention for the From: line in electronic mail clearly varies. There
are three basic notations you see in this line: just an address, such as the one from
[email protected]; an address with the name in parentheses, such as the
message from Linda Wei about halfway down the list; and a line with the person’s
name followed by his or her e-mail address in angle brackets, such as the first listed
line.
Notice the various sites from which I’ve received electronic mail in the past few
days: SFSU.EDU is San Francisco State University, oracle.com is Oracle Corporation
in California, PURDUE.EDU is Purdue University, CompuServe.COM is the CompuServe
network, ccgate.infoworld.com is InfoWorld magazine’s Macintosh network
running cc:Mail, xalt.com is from XALT Corporation, and kent.edu is Kent State
University. The message from [email protected] is from an educational
institution in The Netherlands!
To decode these addresses, you need to have a lot of information, some excellent
guesses, a glance at an Organization: line that might appear in the messages, or, if
you’re on a system with all the latest software, an invocation of the netinfo
command to explain the site you’re curious about.
445
22
Hour 22
446
Sending electronic mail back and forth with users throughout the world is one of the
most exciting and fun parts of learning UNIX. I often read magazine articles, for
example, in which the author lists an electronic mail address. It’s a simple task to zip out a
message if I have questions or kudos on the piece. Many magazines, from the Utne Reader
to MacWorld, even list electronic mail addresses for the editorial staff. Even reporters from
The Wall Street Journal and the New York Times are on the Internet now.
Task 22.2: Talking with Remote Internet Users
To see who is logged in to a remote system, you can use a command called finger,
which by default will show you a summary of who is on the local machine. Add a user
name to the command, and it will show information about the specified account. Specify a
user on a remote system, and you can find out if that user is logged in. Specify just the remote
site, and it shows you who is logged in at the current moment. To check on a local account,
use finger accountname. To make it a remote system, append the hostname: finger
[email protected]. To check all users on a remote site, use finger @host.domain.
Once you’ve ascertained that a friend is logged in to a local or remote system, you can use the
talk program to chat with the person live across the Internet. As opposed to the primitive
line-oriented mode of write, talk is a full-screen program that enables both of you to type
at the same time. Your screen always shows the other person’s typing on the bottom half of
your screen.
1. A quick glance at the output of who shows that there are a lot of people currently
logged in to the local system:
% finger
Login
root
taylor
kippje
adamr
daffnelr
tsa
daffnelr
ben
ben
marteldr
gerlema
mk
mzabel
fritzg
brynta
deckersl
22
Name
TTY Idle
When
root
*co 1:13 Mon 18:02
Dave Taylor
aV
Mon 16:49
Jeff Kip
Ab
Mon 16:41
Adam Coy
*Ae
Mon 18:36
Lawrence Daff
sK
Mon 12:45
Earl the Unctuous Aardva sL
42 Mon 12:48
Lawrence Daff
*sM
Mon 12:49
Ben Moon
sN 8:42 Mon 09:16
Ben Moon
sR 8:47 Mon 09:22
David Martel
*sY
7d Mon 18:41
David Geman
sb
Mon 18:38
Michael Kenzie
*sc 2:48 Mon 08:07
Mary Zabeliski
*sf
Mon 18:45
Geoff Fritzen
sh
9 Mon 18:45
Bryan Ayerson
*si
Mon 18:46
Sharon Deck
sk
3 Mon 18:51
Location
(dov27)
(expert)
(localhost)
(corona)
(corona)
(limbo)
(mac19)
(mk)
(sun1)
(pc43)
(limbo)
(xds31)
447
2. To learn more about the account mk, I can specify that account name to the finger
program:
% finger mk
Login name: mk
(messages off) Real name: Michael Kenzie
Office: Math 204
Home phone:
Directory: /users/mk
Shell: /bin/ksh
Universe: universe(ucb)
Member of groups: utech root actadmin source
On since Dec 13 08:07:12 on ttysc from mk
2 hours 50 minutes Idle Time
No unread mail on this host.
Plan:
You can see that this is full of information. Notice that Michael is currently logged
in to the system (the output says On since Dec 13 08:07:12 on ttysc).
3. To see who is logged in to the USENIX Association main computer in Berkeley,
California, I can use this:
% finger @usenix.org
[usenix.org]
Login
Name
pmui
Peter Mui
ah
Alain Henon
toni
Toni Veglia
diane
Diane DeMartini
mis
Mark Seiden
mis
Mark Seiden
scott
Scott Seebass
lilia
Lilia Carol Scott
mis
Mark Seiden
toni
Toni Veglia
ellie
Ellie Young
scott
Scott Seebass
ellie
Ellie Young
mis
Mark Seiden
TTY
co
Z5
p1
p3
p4
p5
p6
p7
p8
pa
q1
q2
q3
q6
Idle
When
2d Tue 10:10
Mon 15:58
3d Thu 17:09
9 Mon 08:41
3d Thu 21:46
2d Fri 15:18
3d Tue 14:54
1:12 Mon 08:39
3d Thu 22:09
3d Mon 10:38
1:00 Mon 10:33
18: Wed 15:36
1:01 Mon 10:33
1d Fri 11:28
Where
remote # 5408955 Dia
exec
mac2.usenix.ORG
seiden.com
msbnext.internex
biohazard
thing1
seiden.com
exec
boss:0.0
biohazard
boss:0.0
seiden.com
Here you can see that lots of folks are logged in, but that almost everyone has a lot
of idle time. A d suffix indicates the number of days idle. So, you can see that Peter
Mui’s account has been idle for two days.
JUST A MINUTE
To find out what the weather is like in the greater San Francisco area, try
finger [email protected] , which will connect you to the Lawrence
Livermore National Laboratories in Walnut Creek, California.
4. To learn more information about a specific user on a remote site, specify the name
of the user and the name of the user’s system:
% finger [email protected]
[usenix.org]
22
448
Hour 22
Login name: ellie
In real life: Ellie Young
Directory: /staff/ellie
Shell: /bin/csh
On since Dec 13 10:33:57 on ttyq1 from boss:0.0
1 hour 3 minutes Idle Time
Mail last read Mon Dec 13 15:05:08 1993
No Plan.
You can see that Ellie has been logged in since December 13 but has had over an
hour of idle time.
5. To talk with someone on a remote system, use finger to verify that the person is
logged in, not off doing something else (which is what a high idle time usually
suggests), use talk:
% finger [email protected]
Login name: marv
In real life: Marvin Raab
Directory: /u1/marv
Shell: /bin/csh
Logged in since Mon Dec 6 15:22 on ttys8
5 seconds idle time
Mail last read Mon Dec 13 15:22:22 1993
No Plan.
% talk [email protected]
[Waiting for your party to respond]
------------------------------------------------------------------------
On the remote system, here’s what Marvin sees:
Message for marv(ttyaV) from [email protected] at 18:55 ...
talk: connection requested by [email protected].
talk: respond with: “talk [email protected]”
Once he responds, the screen looks like this:
22
[Connected]
_
-----------------------------------------------------------------------Hi Dave! What’s going on?
Notice that the cursor is in the top pane. Anything I enter will be sent along to
Marvin, character by character, so we can interactively chat and even type at the
same time without our words getting jumbled. When I’m done, I simply press ^C
to quit the program.
The finger program offers further information about users on your own and remote
systems, and using it is an essential first step in talking with your friends on the
Internet via talk. Try entering [email protected], and if I’m logged in, what the heck! Try
using talk to say hi interactively!
Task 22.3: Searching Databases with WAIS
The next stop on your tour of Internet information resources is the Wide Area
Information Server, or WAIS. The WAIS system is a collection of databases
accessible through a single search or query, and it was developed as a joint research project
of Apple Computer, Dow Jones News Service, and Thinking Machines, Inc. You can access
WAIS by using telnet to connect to a system called quake.think.com at Thinking Machines,
Inc., in Boston.
WAIS is a database of databases. As of this writing, there are over 500 databases accessible
for conducting searches and queries through the WAIS system. The range of information is
astounding—from databases of acronyms to the CIA World Factbook, from White House
449
22
Hour 22
450
press releases to cold fusion. The program is reasonably friendly to use, though it takes a bit
of experience to be comfortable doing searches.
1. To connect to the WAIS system, use telnet to connect to quake.think.com, and
log in as wais.
You can find an alternative WAIS server at the address wais.com if
quake.think.com is unavailable.
JUST A MINUTE
% telnet quake.think.com
Trying...
Connected to quake.think.com.
Escape character is ‘^]’.
SunOS UNIX (quake.think.com)
login: wais
Last login: Wed Dec 15 11:13:56 from alexia.lis.uiuc.
SunOS Release 4.1.3 (SUN4C-STANDARD) #9: Wed Oct 27 18:18:30 EDT 1993
Welcome to swais.
Please type user identifier (optional, i.e user@host): [email protected]
TERM = (vt100)
Starting swais (this may take a little while)...
After a few seconds, the screen clears and is replaced by this:
SWAIS
Source Selection
Sources: 510
#
Server
Source
Cost
001:
[
archie.au] aarnet-resource-guide
Free
002:
[ndadsb.gsfc.nasa.gov] AAS_jobs
Free
003:
[ndadsb.gsfc.nasa.gov] AAS_meeting
Free
004:
[
munin.ub2.lu.se] academic_email_ conf
Free
005:
[wraith.cs.uow.edu.au] acronyms
Free
006:
[
archive.orst.edu] aeronautics
Free
007:
[ ftp.cs.colorado.edu] aftp-cs-colorado-edu
Free
008:
[nostromo.oes.orst.ed] agricultural-market-news
Free
009:
[
archive.orst.edu] alt.drugs
Free
010:
[
wais.oit.unc.edu] alt.gopher
Free
011:
[
sunsite.unc.edu] alt.sys.sun
Free
012:
[
wais.oit.unc.edu] alt.wais
Free
013:
[alfred.ccs.carleton.] amiga-slip
Free
014:
[
munin.ub2.lu.se] amiga_fish_contents
Free
015:
[
coombs.anu.edu.au] ANU-Aboriginal-EconPolicies
$0.00/minute
016:
[
coombs.anu.edu.au] ANU-Aboriginal-Studies
$0.00/minute
017:
[
coombs.anu.edu.au] ANU-Ancient-DNA-L
$0.00/minute
018:
[
coombs.anu.edu.au] ANU-Ancient-DNA-Studies
$0.00/minute
Keywords:
<space> selects, w for keywords, arrows move, <return> searches, q quits, or ? _
22
You can see here a table of contents of the different databases available for searching through WAIS. Although they’re all free of charge at this point, it is entirely
possible that at some point in the future some of these databases will have costs
associated with them.
Type J and K to move up and down a screen at a time, respectively, and j and k to
move up and down a single source. To add a database to a search, type . to select
it. An asterisk will show up just before the name of the system that holds the
database selected. Type q to quit.
2. To find out more about a particular database, use the v command for version
information. When I choose that option for acronyms, here’s what I find out:
Name:
acronyms.src
Directory:
/sources/
Maintainer:
[email protected]
Selected:
Yes
Cost:
Free
Server:
wraith.cs.uow.edu.au (Accessed)
Service:
210
Database:
acronyms
Description:
Server created with WAIS release 8 b5 on Oct 23 10:49:48 1992 by
[email protected]
A public domain database of acronyms and abbreviations maintained
by Dave Sill ([email protected]).
The files of type one_line used in the index were:
/shr/lib/wais/wais-sources/acronyms
(END)
3. I’m interested in how many acronyms have the word mail in them, so I choose the
acronym database by pressing the spacebar (an asterisk appears to show that the
database has been selected). Then I use w to specify a keyword, in this case mail. A
quick search, and the screen now looks like this:
SWAIS
#
001:
002:
003:
004:
005:
006:
007:
008:
009:
010:
011:
012:
013:
Score
[1000]
[ 333]
[ 333]
[ 333]
[ 333]
[ 333]
[ 333]
[ 333]
[ 333]
[ 333]
[ 333]
[ 333]
[ 333]
(
(
(
(
(
(
(
(
(
(
(
(
(
Source
acronyms)
acronyms)
acronyms)
acronyms)
acronyms)
acronyms)
acronyms)
acronyms)
acronyms)
acronyms)
acronyms)
acronyms)
acronyms)
Search Results
Item
Title
Lines
EMAIL - Electronic MAIL, “E-MAIL”
1
ECOM
- Electronic Computer Originated Ma
1
EMA
- Electronic Mail Association
1
IMAP
- Interactive Mail Access Protocol 1
IMAP3 - Interactive Mail Access Protocol
1
MIME
- Multipurpose Internet Mail Extens 1
MO
- Mail Order
1
MTA
- Mail Transfer Agent
1
MUA
- Mail User Agent
1
MX
- Mail eXchange
1
PBM
- Play By Mail game
1
PEM
- Privacy Enhanced Mail 1
RMS
- Royal Mail Ship
1
451
22
Hour 22
452
014:
015:
016:
[ 333] (
[ 333] (
[ 333] (
acronyms)
acronyms)
acronyms)
SMTP
USM
VMS
- Simple Mail Transfer Protocol
- United States Mail
1
- Voice Mail System
1
1
<space> selects, arrows move, w for keywords, s for sources, ? for help _
4. Here’s another way this can be helpful: WAIS lists a database of recipes, and I’ve
been looking for a good oatmeal cookie recipe for quite a while. I can search the
recipe database by returning to the main WAIS screen and then entering /rec to
move to that particular database:
SWAIS
#
397:
398:
399:
400:
401:
402:
403:
404:
405:
406:
407:
408:
409:
410:
411:
412:
413:
414:
Server
[
munin.ub2.lu.se]
[ wais.wu-wien.ac.at]
[
wais.oit.unc.edu]
[
wais.oit.unc.edu]
[bloch.informatik.uni]
[
gopher.uv.es]
[
wais.cic.net]
[
ds.internic.net]
[
ns.ripe.net]
[
ns.ripe.net]
[
ns.ripe.net]
[ cmns-moon.think.com]
[
wais.cic.net]
[
mpcc3.rpms.ac.uk]
[ cmns-moon.think.com]
[
uniwa.uwa.oz.au]
[athena3.cent.saitama]
[RANGERSMITH.SDSC.EDU]
Source Selection
Source
rec.gardens
rec.music.early
rec.pets
recipes
reports-abstracts
Research-in-Surgery
rfc-index
rfcs
ripe-database
ripe-internet-drafts
ripe-rfc
risks-digest
roget-thesaurus
RPMS-pathology
RSInetwork
s-archive
saitama-jp
Salk_Genome_Center
Sources: 510
Cost
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
Free
<space> selects, w for keywords, arrows move, <return> searches, q quits, or ?
CAUTION
_
One problem I consistently have with the WAIS programs is that I can’t
use my Backspace key to erase previous search words. The trick to getting
around this is to use û to erase the entire line of keywords!
I choose the recipes database, again by pressing the spacebar, then search for
keywords cookie and oatmeal to see what kind of oatmeal cookie recipes are
available. It indicates 18 matches, sorted in order of the “quality” of the match—
the more each keyword occurs, the better the hit. The first number in square
22
brackets indicates the “quality” of the match, with 1,000 being the best possible
score. This time, however, the recipe I’m looking for appears to be the lowest-rated
in the list:
SWAIS
#
001:
002:
003:
004:
005:
006:
007:
008:
009:
010:
011:
012:
013:
014:
015:
016:
017:
018:
Score
[1000]
[ 957]
[ 696]
[ 522]
[ 522]
[ 522]
[ 435]
[ 391]
[ 348]
[ 348]
[ 348]
[ 348]
[ 348]
[ 304]
[ 304]
[ 261]
[ 261]
[ 261]
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
(
Source
recipes)
recipes)
recipes)
recipes)
recipes)
recipes)
recipes)
recipes)
recipes)
recipes)
recipes)
recipes)
recipes)
recipes)
recipes)
recipes)
recipes)
recipes)
Search Results
Item
Title
Lines
shafer@rig Re: COLLECTION BAKERY Pumpkin
186
shafer@rig Re: COLLECTION BAKERY VEG Bis 1038
Anne Louis Re: BREAD: Bread Recipes Coll 1101
darsie@eec Re: Re: world-wide cookie rec
130
kyoung@prs Re: Diabetic Cookie Recipes
91
the1edr@ca Re: 3 Recipes for Oatmeal Pea
122
WHITEJER@c Re: Re: REQUEST Cookie dough
53
julie@eddi Re: Re: REQUEST Cookie dough
47
anne@csrux Re: Re: REQUEST Cookie dough
53
arielle@ta Re: Appetizers (Long)
1591
kyoung@prs Re: Diabetic Cookie Recipes C
77
arielle@ta Re: Muffins 5
717
arielle@ta Re: Appetizers
1590
springer@k Re: Re: REQUEST Cookie dough
44
kyoung@prs Re: Diabetic Treats Cont’d.
109
arielle@ta Re: RECIPE: Ice Cream Sandwic
63
arielle@ta Re: REQUEST Cookie dough for
22
laura@hobb Re: Oatmeal Rasin Cookies (ca
31
<space> selects, arrows move, w for keywords, s for sources, ? for help _
To read recipe number 18, I can enter 18 to retrieve the recipe itself:
Getting “laura@hobb Re: Oatmeal Rasin Cookies (cake like)” from recipes.src...
Newsgroups: rec.food.recipes
From: [email protected] (Laura Smith)
Subject: Oatmeal Rasin Cookies (cake like)
Apparently-To: [email protected]
Organization: The Mystery University
Date: Tue, 22 Dec 1992 15:33:33 GMT
Approved: [email protected]
Lines: 18
Hi,
My father was over helping me bake cookies this year for christmas.
He got to talking about an oatmeal cookie that his mother used to make.
These cookies were almost like little individual oatmeal rasin cakes
Unfortunately.. he never got the recipe written down.... I’d really like to
find a recipe and surprise him with them.
--More-- _
453
22
Hour 22
454
Voilà! I can sit at my UNIX system and dig up just about anything on the Internet,
even cookie recipes!
Of the different services on the Internet, the WAIS system is the one I find the most
promising, yet least useful. There are lots of problems with the system, but it’s
evolving at such a fast pace that I encourage you to try it for yourself. By the time you read
this, the program undoubtedly will have changed a fair bit.
Task 22.4: Having the Whole World with gopher
At this point, you’re probably wondering how people are actually supposed to choose
from and navigate all these different services—and rightfully so! A team of programmers at the University of Minnesota wondered just that. They realized that what they wanted
was a “go-for,” a program that would “go for things.” Conveniently the gopher is their school
mascot, so the gopher program was born.
Of all the different systems on the Internet, gopher is undoubtedly the easiest to use. It has
a simple, menu-based interface that enables you to step through information sources,
seamlessly switching from machine to machine throughout the Internet. The program offers
some helpful customization, too: As you travel through what’s called gopherspace, you can
mark interesting locations with a bookmark (simply press a at the item), then zoom straight
to your list of bookmarks with the v key, to view your bookmarks. If you have bookmarks,
gopher -b will start you up with your bookmark page; otherwise, it will display the default
gopher introductory page, which varies quite a bit from site to site.
TIME SAVER
If you don’t have the gopher program on your system, you can log in as
gopher at consultant.micro.umn.edu, gopher.uiuc.edu , or
panda.uiowa.edu . Use telnet to connect.
1. I enter gopher at the command line of my account at the UTech University, and
the screen is rewritten:
Internet Gopher Information Client v1.12S
Root gopher server: thorplus.utech.edu
-->
22
1.
2.
3.
About UTech University.
About THOR+ the UTech University Libraries Gopher Site.
Other Information Servers at UTech University/
455
4. Other information Servers on the Internet/
5. Thor+ Suggestion Form <TEL>
6. Administrivia/
7. E-Mail & Telephone Directory for Utech & World Wide/
8. Library Catalogs and Gophers/
9. University Libraries/
10. Weather Reports and Maps/
11. Interesting items on the Net (12/7/93)/
12. Current Contents On Diskette/
13. Instructions for searching Directories of all UTech Gophers .
14. Search Directories of all UTech Gophers (experimental) <?>
15. *************Explore Internet Teleconference********/
Press ? for Help, q to Quit, u to go up a menu
Page: 1/1
2. By contrast, if I were logged in to the Whole Earth ‘lectronic Link (well) computer
in San Francisco, I’d get a completely different first screen:
Internet Gopher Information Client v1.11
Root gopher server: gopher2.tc.umn.edu
-->
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
Information About Gopher/
Computer Information/
Discussion Groups/
Fun & Games/
Internet file server (ftp) sites/
Libraries/
News/
Other Gopher and Information Servers/
Phone Books/
Search Gopher Titles at the University of Minnesota <?>
Search lots of places at the University of Minnesota <?>
University of Minnesota Campus Information/
Page: 1/1
3. The sixth entry—Libraries/—sounds interesting, and because it ends with a slash,
I can tell that it will move me to another set of menu choices in gopher. The two
lines that end with <?> will actually invoke a program (probably to connect me to
the University of Minnesota), and lines that end with a dot are files and can be
viewed by choosing them.
To move to a specific location, I can enter its number or use j and k to move up
and down, just like in vi. Pressing Return chooses the specific item, so I press j five
times to move down five items (the arrow moves so that it points to item 6), then I
press Return, which changes the screen:
22
Hour 22
456
Libraries
-->
1.
2.
3.
4.
5.
6.
7.
Electronic Books/
Electronic Journal collection from CICnet/
Information from the U.S. Federal Government/
Library Catalogs via Telnet/
Library of Congress Records/
Newspapers, Magazines, and Newsletters /
Reference Works/
Page: 1/1
4. Electronic books sound interesting, but reference works could be even more
interesting, so I type 7, which instantly moves the arrow to the last item, and then I
press Return:
Reference Works
-->
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
ACM SIGGRAPH Online Bibliography Project/
American English Dictionary (from the UK) <?>
CIA World Fact Book 1991/
Current Contents/
ERIC-archive.
ERIC-archive Search <?>
Periodic Table of Elements/
Roget’s Thesaurus (Published 1911)/
The Hacker’s Dictionary/
U.S. Geographic Names Database/
U.S. Telephone Area Codes/
US-State-Department-Travel-Advisories/
Webster’s Dictionary/
Page: 1/1
5. The entry for U.S. State Department travel advisories looks valuable, so I’ll add it
to my bookmark collection:
22
457
7. Periodic Table of Elements/
***********************************************************************
*
*
* Name for this bookmark?
US-State-Department-Travel-Advisories
*
*
*
*
[Cancel ^G] [Accept - Enter] *
*
*
***********************************************************************
The default name works fine, so I press Return to move into the choice, then I
see three choices: 1. Search US-State-Department-Travel-Advisories <?>, 2.
Current-Advisories/ , and 3. FTP-Archive/ . I opt to see what’s current, and
type 2:
Current-Advisories
-->
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
afghanistan.
albania.
algeria.
andorra.
angola.
antigua-&-barbuda.
argentina.
armenia.
australia.
austria.
azerbaijan.
bahamas.
bahrain.
bangladesh.
barbados.
belarus.
belgium.
belize.
Page: 1/13
Notice that this time the bottom-right corner indicates that this is page 1 of 13, so
there’s a lot more information. To move to the next page of information, use the +
key; to move to the previous page, use -. I have visited Belize, so I’d be interested
to see whether there are any current travel advisories on the country. I choose 18,
and the following information is displayed on my screen:
22
Hour 22
458
STATE DEPARTMENT TRAVEL INFORMATION - Belize
============================================================
Belize - Consular Information Sheet May 27, 1993
Country Description: Belize is a developing country. Its tourism facilities
vary in quality.
Entry Requirements: A passport, a return/onward ticket, and sufficient funds
are required for travel to Belize. U. S. citizens who stay less than three
months do not need visas. However, for visits exceeding one month, travelers
must obtain permits from immigration authorities in Belize. For further
information, travelers may contact the Embassy of Belize at 2535 Massachusetts
Avenue N.W., Washington, D.C. 20008, tel. (202) 332-9636, the Belize Consulate
in Miami, or the Belize Mission to the U.N. in New York.
Medical Facilities: Medical care is limited. Doctors and hospitals often
expect immediate cash payment for health services.
U.S. medical insurance is not always valid outside the United States. In some
cases, supplemental medical insurance with specific overseas coverage has
proved useful. For additional health
--More--(9%)[Hit space to continue, Del to abort] _
This is only nine percent of the information, so there’s a lot more to view. Fortunately, I can electronically mail this file to myself when I’ve finished viewing it, by
selecting the m, or mail file, command at the end-of-listing prompt.
JUST A MINUTE
Gopher notation usually has a nested series of lines indicating the actual
text that you’d find on a line of gopher output, so this search would be
written much more succinctly as:
Libraries/
Reference Works/
US-State-Department-Travel-Advisories/
Current Advisories
belize.
I quit by typing q at any prompt in the gopher system.
7. Back at the university, I have heard that there are electronic books available via
gopher. I am particularly interested in Paradise Lost, a book I read years ago. After a
bit of nosing about, I found it through the following gopherspace path:
Other Information Servers on the Internet/
Academic Resources on the Internet (by Subject)/
Electronic Journals & Texts/
Project Gutenberg: Clearinghouse for Machine Readable Texts/
etext92/
AAINDEX.NEW.
22
The AAINDEX.NEW file produced a list of books available through this clearinghouse
for electronic books. Skipping the introductory matter, I find that the following
books are currently online:
(Books from earlier years will available in 1992)
(but not yet: to be announced, don’t ask yet!!!)
1971 Declaration-Independence
(whenxxxx.xxx)
1972 Bill of Rights
(billxxxx.xxx)
1973 U.S. Constitution
(constxxx.xxx)
1974-1982 The Bible
(biblexxx.xxx)
1983-1990 Complete Shakespeare
(shakesxx.xxx)
(Watch for these entries to be moved below later.
The Bible mentioned above is a different edition from the one we just post
for Easter, 1992)
Books currently available on mrcnext (do a dir):
(These 1991 etexts are now in> cd /etext/etext91)
Jan 1991 Alice in Wonderland
(alice29x.xxx)
Feb 1991 Through the Looking Glass
(lglass16.xxx)
Mar 1991 The Hunting of the Snark
(snark12x.xxx)
Apr 1991 1990 CIA World Factbook
(world11x.xxx)
May 1991 Moby Dick (From OBI)*
(mobyxxxx.xxx)
Jun 1991 Peter Pan (for US only)**
(peter14a.xxx)
Jul 1991 The Book of Mormon
(mormon11.xxx)
Aug 1991 The Federalist Papers
(feder11x.xxx)
Sep 1991 The Song of Hiawatha
(hisong10.xxx)
Oct 1991 Paradise Lost
(plboss10.xxx)
Nov 1991 Aesop’s Fables
(aesop10x.xxx)
Dec 1991 Roget’s Thesaurus
(roget11x.xxx)
*Moby Dick is missing Hour 72
**Please do not download Peter Pan outside the US
(These 1992 etext releases in> cd /etext/etext92)
Jan 1992 Frederick Douglass
(duglas10.xxx)
Jan 1992 O Pioneers! Willa Cather
(opion10x.xxx)
Feb 1992 1991 CIA World Factbook
(world91a.xxx)
Feb 1992 Paradise Lost (Raben)
(plrabn10.xxx)
Mar 1992 Far From the Madding Crowd
(crowd13x.xxx)
Mar 1992 Aesop’s Fables (Advantage)
(aesopa10.xxx)
Apr 1992 Data From the 1990 Census
(uscen901.xxx)
Apr 1992 New Etext of Bible (KJV)
(bible10x.xxx)
May 1992 Sophocles’ Oedipus Trilogy
(oedip10x.xxx)
May 1992*Herland (not yet in place
(hrlnd10x.xxx)
There are more resources than this, but here you can see that if you’re interested in
obtaining a copy of Paradise Lost, you can use gopher to find the book and have it
sent via Internet to your home account—in a matter of a few steps!
The gopher system offers a wide variety of capabilities, as you can see, and the
connectivity is astounding. One aid to finding information in gopherspace is a search
program called veronica, with which you can specify one or more words that you think might
show up in the one-line menu listings. Overall, I find it enjoyable just to wander about and
459
22
Hour 22
460
see what’s available. At any point, you can type u to return to a previous menu, so you can
wander to your heart’s content.
I’m not making this up: veronica stands for very easy rodent-oriented
net-wise index to computerized archives.
JUST A MINUTE
Task 22.5: Visiting Libraries Around the World
As it turns out, I’ve written another book, one called Global Software. How about
joining me as I travel through the Internet to various libraries to see what universities
have my book?
1. The first library I visit is the National Library in Venezuela (Biblioteca Nacional).
With gopher, the library computer is only seven steps away from the very top!
Libraries/
Library Catalogs via Telnet/
Library Catalogs from Other Institutions/
Catalogs Listed by Location/
Americas/
Venezuela/
Biblioteca Nacional <TEL>
Once I log in as biblio, the screen looks like this:
Catalogo Bib. Nacional
Introduccion
-----------------------------------------------------------------------------Bienvenido al Catalogo Automatizado
S A I B I N
Use los siguientes comandos:
Para buscar por:
A=
Autor
T=
Titulo
M=
Materia
K=
Palabra clave
C=
Cota
Ud. puede iniciar una busqueda desde cualquier pantalla
Para mayor informacion de busqueda en el Catalogo, presione <ENTER>.
Para ver informacion sobre las BASES DE DATOS, escriba NOTI y
presione <ENTER>.
Ademas de LUIN, ud. puede usar el comando LUC2 donde encontrara
22
-----------------------------------------------INGRESE COMANDO BUSQUEDA
NOTicia
+ Pag. 1 de 4 -------<F8> AVAnza pag.
Prox. Comando: _
Fortunately, my Spanish is sufficient to figure this out.… I use t= to search by title
(the command is T=GLOBAL SOFTWARE) for my book Global Software.
Solicitud de Busqueda: T=GLOBAL SOFTWARE
Catalogo Bib. Nacional
Resultados de Busqueda: 0 Entradas Encontadas
No existen Entradas por Titulo
-----------------------------------------------------------------------------No se encontraron Entradas de Titulo
Las posibles razones para este mensaje son:
1.
2.
3.
Material no esta en la base de datos (Busque en el fichero.)
Material no pertenece a la Biblioteca (Consulte al personal de referencia)
Comando o termino(s)
incorrecto. (Pruebe con otro comando o cambie el
termino de busqueda.)
Verifique en la busqueda lo siguiente:
--Asegurese de que estan correctamente escritos. Si no esta seguro del titulo completo o de como se escribe acortelo al final.
--Omita todos los articulos (a, en, el).
--Elimine los signos de puntuacion.
Recuerde: Ud. puede revisar su busqueda editandola en la linea de comandos).
------------------------------------------------ Pag. 1 de 1 -------------COMenzar
REVisa busquedas realizadas
OTRas opciones
Prox. Comando: _
Ay caramba! Mi libro no está en al biblioteca nationál. Qué es la vida!
That is, there isn’t a copy of my book in the library. Such is life. I can use salir (to
leave) to log out and return to gopher.
2. My next library to visit is to Australia. To get there, I need to back up a few levels
in gopher and travel down a different path: Asia and Pacific/ lead to Australia/,
where I can choose the Queensland University of Technology/:
461
22
Hour 22
462
Q U T
L A T E S T
N E W S
L I B R A R Y
EXCELLENCE IN INFORMATION SERVICES
***********************************
*
OPTIONS
*
***********************************
*
*
* 1.
Library Catalogue
*
* 2.
Library Opening Hours
*
* 3.
Logoff
*
*
*
*
*
***********************************
or press <HELP>, then <RETURN>
This looks likely, so I choose 1:
L I B R A R Y
C A T A L O G U E
****************************************************************************
*
MAIN MENU
*
****************************************************************************
1. Titles
2. Personal Authors
3. Subjects
4. Series Titles
5. Corporate Authors & Conferences
6. Keyword search across files 1-5
7. Call Numbers
8. Publishers
9. Limited Access Collection - Subject
10. Limited Access Collection - Lecturer
****************************************************************************
*
3 most important keys are:
*
*
<ENTER> to activate all commands
*
*
<HELP> or <shift and ?> then <ENTER> for advice
*
*
Full stop <.> then <ENTER> to backtrack or terminate a search.*
****************************************************************************
Please select _
I want to select by title, so again I type 1:
22
463
S E A R C H
T E X T
E N T R Y
************************************************************************
TITLE ALPHABETIC BROWSING
************************************************************************
Please type in some text that you think may occur at the start of a
title of a work that you are attempting to find.
For example, you could type
ECONOMICS AND DEMOGRAPHY
ECONOMICS AND
ECON
or
or
And remember to use the help key if you need more information.
Please enter your search text...
************************************************************************
> global software
I press Return to get the news:
B R O W S E
H E A D I N G
S E L E C T I O N
Search text: GLOBAL SOFTWARE
Headings retrieved: 100
More pages
** TITLES **************************************************************
No. Works
1.
1
Global simulation models : a comparative study
==> GLOBAL SOFTWARE
2.
1
Global solutions : innovative approaches to world
problems: selections from The Futurist
3.
1
Global sourcing strategy : R&D, manufacturing, and
marketing interfaces
4.
1
Global stakes : the future of high technology in America
5.
2
Global status of mangrove ecosystems
6.
1
Global stock market reforms
7.
1
Global Strategic Management : Impact on New Frontiers....
8.
1
Global strategic management perspectives
************************************************************************
Next Page
Gather Headings
Keyword Search
Previous Page
List Chosen Headings
Top of List
SELECT _
Nope, it’s not there either, but the book Global Status of Mangrove Ecosystems
sounds quite interesting. Next time I’m on this computer system, I should look up
the reference to learn more about mangrove ecosystems.
22
Hour 22
464
3. I’ll try one more university before I give up hope! I’ll check the various libraries of
the University of California. Again, the process is to step back in the gopher tree
and select Americas/, United States/, California/, University of California
(MELVYL) <TEL>/ :
Welcome to the University of California’s
MELVYL* LIBRARY SYSTEM
-------------------- =>> SYSTEM NEWS <<= ------------------------
------------------------------------------------------------------
(c)1984. *Registered TM of The Regents of the University of California.
========================================================================
OPTIONS:
Choose option, or type command to enter the CATALOG database.
HELP
- For help in getting started.
[return]
- Press RETURN to choose a database for searching.
START <db>
- Type START <db name> to begin searching in a database.
-> only _
First, I can use the shortcut of entering START
University of California library holdings:
CAT
to start with a catalog of all
Welcome to the MELVYL CATALOG Database
Contents:
As of 12/15/93, approximately 7,563,498 titles
representing 11,761,300 holdings for materials in the
University ofCalifornia libraries and the California
State Library.
Coverage:
All publication dates but incomplete for some libraries.
--=>> NEWS <<=--
---------------------------------------------------------------------OPTIONS:
Type an option and press RETURN, or type any command.
HELP
22
-
For help in getting started.
E GUIDE
START
END
CAT-> _
-
For a brief guide to using the Catalog database.
-
To start over or change databases.
To end your session.
Did you see the number at the top of that screen? This database lists over 7.5
million different books, representing over 11 million holdings. That’s quite
impressive.
4. To save time again, I use the find command, specifying a title word: FIND TW
GLOBAL SOFTWARE , which results in nine matches:
CAT-> f tw global software
Search request: F TW GLOBAL SOFTWARE
Search result: 9 records at all libraries
Type D to display results, or type HELP.
CAT-> _
Typing D displays the first page of matches:
Type HELP for other display options.
1. Clapes, Anthony Lawrence.
Softwars : legal battles for control of global software industry /
Anthony Lawrence Clapes. Westport, Conn. : Quorum Books, 1993.
HAST 5th Stks K1443.C6 C56 1993
UCB
Bus&Econ K1443.C6 C56 1993
UCB
Law Lib
K89 .C48
UCI
Main Lib K1443.C6 C56 1993
UCLA College
K 1443 C6 C56 1993
UCSC McHenry
K1443.C6C56 1993
UCSD Central
K1443.C6 C56 1993
2. A computer software system for the generation of global ocean tides
including self-gravitation and crustal loading effects,
by ronald h estes.1977.
UCSD Scripps
FICHE XSX 1 N77-23709 Floor 1 Microform
Press RETURN to see the next screen.
CAT->
465
22
Hour 22
466
I step forward a page or two:
Type HELP for other display options.
5. Taylor, Dave, 1962Global software : developing applications for the international
market, Dave Taylor. New York : Springer-Verlag, c1992.
UCB
Engin
QA76.76.D47 T39 1992
UCI
Main Lib QA76.76.D47 T39 1992
UCSC Science
QA76.76.D47T39 1992
6. United States. General Accounting Office.
Air Force Global Weather Central initiates positive action to
assess adequacy of software inventory : report to the Secretary
of the Air Force by the U.S. General Accounting Office. Washington,
D.C. : The Office, [1983].
UCR
Rivera
GA 1.13:IMTEC-84-4 Govt.Pub Microfiche US
UCSD Central
GA 1.13:IMTEC-84-4 Documents Fiche
CSL
Main Lib GA 1.13:IMTEC-84-4 Govt Pubs
Press RETURN to see next screen. Type PS to see previous screen.
CAT-> _
Aha! You can see that there are copies of my book at the University of California at
Berkeley (UCB), at Irvine (UCI), and at Santa Cruz (UCSC).
There are hundreds, if not thousands, of libraries connected to the Internet. If a
reference book exists, you should be able to find a reference citation.
Task 22.6: All the News That’s Fit or Otherwise
No discussion of the Internet would be complete without a brief foray into the
largest, most active, and most varied discussion forum in the world—the Usenet.
Imagine a bulletin board on the wall. Imagine that as people pass it, they glance at what’s
there, and if they have something to add, they stick their note up, too. Now (and here’s the
big leap), imagine that there are thousands of bulletin boards in this building, and that there
are actually tens of thousands of buildings throughout the world, each with its own
“identical” copy of the bulletin boards. Got it? That’s Usenet.
Usenet was created in 1979, when two graduate students at Duke University, Tom Truscott
and Jim Ellis, hooked their computer to another computer at the University of North
Carolina. In 1980, there were two sites with Usenet. Today, at the very end of 1993, there
are an estimated 120,000 sites on Usenet, representing over 4.2 million participants.
22
A true experiment in free speech and barely controlled anarchy, the range of discussions,
called newsgroups, is astonishing. It covers everything from computer modem protocols
(comp.dcom.modem) to Macintosh programming (comp.sys.mac.programmer), topics of relevance to single men and women (soc.singles), abortion (talk.abortion) to the wonderful
TV show “Mystery Science Theater 3000” (alt.tv.mst3k). Whatever your interest, there’s
a group on the Net that talks about just what you’re thinking about!
JUST A MINUTE
A protocol is a language that different systems use to speak to each other
so that they can interoperate. A modem protocol is the language that your
modem uses to interact with your computer. Sound exciting?
The difficulty with Usenet is that the majority of tools designed to help read the volumes of
information actually do precious little to help. The first puzzle is to find the groups that you’d
like to read, and although almost all Usenet sites have a succinct database of what each group
discusses, little Usenet software actually knows about it. Here is a very simple C shell alias that
will help out:
alias findgroup _grep -i \!* /usr/local/lib/news/newsgroups_
At your site, this file might be available also as /usr/lib/news/newsgroups. Newsgroups are
organized into seven primary hierarchies: comp groups are computer and programming
related, sci groups discuss scientific issues, misc groups cover miscellaneous topics, rec are
recreational, talk groups are for controversial and often heated discussion groups, soc are
social groups, and news are groups containing news of the world or at least news of the network
itself.
One final hierarchy worth mentioning is the alt.* set of groups that are the spot for semidisorganized anarchy on the Net: Essentially anyone can create an alt group with ease, so,
as you might expect, these groups are the most varied. Some examples are the excellent
alt.activism for political activists and alt.books.technical for discussion of computer
books (like this very book!), juvenile groups such as alt.binaries.sounds-armpit.noises
for, one presumes, audio files that contain sounds of armpit noises, and alt.elvis.sighting
for those elusive sightings of The King. If you can’t quite tear yourself away from your video
game system long enough even to eat, perhaps reading alt.get.a.life.nintendo.addicts
will help.
One problem with the alt groups is that it’s much more difficult to find out what’s out there.
Because there is considerably less control and organization, the convenient one-line descriptions in the newsgroups file don’t contain descriptions of these alternative groups. Really, the
best solution is to search through your .newsrc itself for key words or abbreviations. To find
alt groups that discuss Disney, for example, I could try grep disney .newsrc | grep _âlt_
to find alt.fan.disney.afternoon.
467
22
Hour 22
468
Once you’ve found a group to read, it’s time to choose from among the many possible
packages. Perhaps the most popular is rn, or read news, written by Larry Wall. Another
alternative, nn, offers a more screen-oriented view from Kim Storm, and a third possibility,
patterned after the Elm Mail System, is tin, designed and written by Iain Lea.
JUST A MINUTE
Of the many programs available for reading netnews, I prefer tin, but
because rn is so prevalent, I will use it in this hour. I nonetheless encourage you to use a local copy of tin if it’s available. Remember also that
you always can use archie to find a local copy. (I talk about archie in
the next hour.)
The rn program not only has more options than you can shake a stick at, but it actually has
more options than even a tree full of sticks could cover! Table 22.2 lists a small number of
its particularly useful options.
Table 22.2. The most useful rn starting options.
Option
Meaning
-/
Set SAVEDIR, so that articles you save are stored in a subdirectory of
~/News named after the group, with the article name corresponding to its
numeric identifier on the system.
Check for news and indicate if any has arrived.
Make each page of an article start at the top of the screen.
Suppress the header hdr in news articles.
Leave information on screen as long as possible.
Force mailbox format for all saved files.
Use inverse video for highlighted information.
Force normal, non-mailbox format for all saved files.
Restart within the last newsgroup read during the previous session.
Use subject search mode when possible.
-c
-e
-hhdr
-L
-M
-m
-N
-r
-S
That’s a lot of options! My alias for starting up the rn program is rn -L -M
-m -e -S -/ .
JUST A MINUTE
22
Did I mention the RNINIT environment variable yet? In addition to using starting options,
the rn program also can read a variety of different options from environment variables.
Indeed, any option that is specified at the command line also can be specified in the RNINIT
variable. As a result, my RNINIT is this:
% echo $RNINIT
-hmessage -hreference -hdate-r -hsender -hsummary -hreply -hdistr -hlines
-hline -hfollow -hnews -hkey -hresent -hreturn -hto -hx-original -hx-sun
-hx-note -horiginator -hnntp
This causes the program to suppress the display of all the specified headers in individual news
articles. There are just so many options that it’s overwhelming. Let’s go into the program and
see what it looks like!
1. First, I use the alias findgroup to identify a few newsgroups I’d like to read:
% findgroup mac
biz.dec.ip
IP networking on DEC machines.
vmsnet.internals
VMS internals, MACRO-32, Bliss, gatewayed to MACRO32 list.
gnu.emacs.announce Announcements about GNU Emacs. (Moderated)
gnu.emacs.bug
GNU Emacs bug reports and suggested fixes. (Moderated)
gnu.emacs.gnews
News reading under GNU Emacs using Weemba’s Gnews.
gnu.emacs.gnus
News reading under GNU Emacs using GNUS (in English).
gnu.emacs.help
User queries and answers.
gnu.emacs.sources ONLY (please!) C and Lisp source code for GNU Emacs.
gnu.emacs.vm.bug
Bug reports on the Emacs VM mail package.
gnu.emacs.vm.info Information about the Emacs VM mail package.
gnu.emacs.vms
VMS port of GNU Emacs.
gnu.epoch.misc
The Epoch X11 extensions to Emacs.
comp.binaries.acorn Binary-only postings for Acorn machines. (Moderated)
comp.binaries.mac Encoded Macintosh programs in binary. (Moderated)
comp.emacs
EMACS editors of different flavors.
comp.lang.forth.mac
The CSI MacForth programming environment.
comp.lang.lisp.mcl
Discussing Apple’s Macintosh Common Lisp.
comp.org.acm
Topics about the Association for Computing Machinery.
comp.os.mach
The MACH OS from CMU & other places.
comp.os.msdos.misc
Miscellaneous topics about MS-DOS machines.
comp.os.msdos.programmer
Programming MS-DOS machines.
comp.os.os2.programmer Programming OS/2 machines.
comp.sources.mac
Software for the Apple Macintosh. (Moderated)
comp.sys.mac.advocacy
The Macintosh computer family compared to others.
comp.sys.mac.announce
Important notices for Macintosh users. (Moderated)
comp.sys.mac.apps
Discussions of Macintosh applications.
comp.sys.mac.comm
Discussion of Macintosh communications.
comp.sys.mac.databases Database systems for the Apple Macintosh.
comp.sys.mac.digest
Apple Macintosh: info&uses, no programs. (Moderated)
comp.sys.mac.games
Discussions of games on the Macintosh.
comp.sys.mac.hardware
Macintosh hardware issues & discussions.
comp.sys.mac.hypercard The Macintosh Hypercard: info & uses.
comp.sys.mac.misc
General discussions about the Apple Macintosh.
469
22
Hour 22
470
comp.sys.mac.oop.macapp3 Version 3 of the MacApp object oriented system.
comp.sys.mac.oop.misc
Object oriented programming issues on the Mac.
comp.sys.mac.programmer Discussion by people programming the Apple Macintosh.
comp.sys.mac.system
Discussions of Macintosh system software.
comp.sys.mac.wanted
Postings of “I want XYZ for my Mac.”
comp.sys.sgi.graphics
Graphics packages and issues on SGI machines.
comp.sys.sgi.misc
General discussion about Silicon Graphics’s machines.
comp.text.tex
Discussion about the TeX and LaTeX systems & macros.
comp.unix.aux
The version of UNIX for Apple Macintosh II computers.
misc.forsale.computers.mac
Apple Macintosh related computer items.
rec.games.diplomacy
The conquest game Diplomacy.
sci.nanotech
Self-reproducing molecular-scale machines. (Moderated)
Remember that these are only groups that have the word mac in them. The group
comp.sys.mac.announce sounds like it might be interesting. Now, to find a few
more:
% findgroup writing
comp.edu.composition Writing instruction in computer-based classrooms.
misc.writing
Discussion of writing in all of its forms.
% findgroup education
k12.ed.art
Art curriculum in K-12 education.
k12.ed.business
Business education curriculum in grades K-12.
k12.ed.health-pe Health and Physical Education curriculum in grades K-12.
k12.ed.life-skills
Home Economics and Career education in grades K-12.
k12.ed.math
Mathematics curriculum in K-12 education.
k12.ed.music
Music and Performing Arts curriculum in K-12 education.
k12.ed.science
Science curriculum in K-12 education.
k12.ed.soc-studies Social Studies and History curriculum in K-12 education.
k12.ed.special
K-12 education for students w/ handicaps or special needs.
k12.ed.tag
K-12 education for talented and gifted students.
k12.ed.tech
Industrial Arts and vocational education in grades K-12.
k12.lang.art
Language Arts curriculum in K-12 education.
comp.ai.edu
Applications of Artificial Intelligence to Education.
comp.edu
Computer science education.
misc.education
Discussion of the educational system.
sci.edu
The science of education.
% findgroup movies
rec.arts.movies
Discussions of movies and movie making.
rec.arts.movies.reviews Reviews of movies. (Moderated)
rec.arts.sf.movies
Discussing SF motion pictures.
rec.arts.startrek.current
New Star Trek shows, movies and books.
% findgroup film
rec.arts.startrek.reviews
Reviews of Star Trek books, episodes, films, &c
(Moderated)
Now I have a list of groups to check out: comp.sys.mac.announce, misc.writing,
sci.edu, and rec.arts.movies .
2. It’s time to start up the rn program so I can read these groups:
% rn
*** NEWS NEWS ***
Welcome to version 4.4 of rn (patch level 4). This version corrects
many bugs of the previous version and has some enhancements you may
find useful. Type “man rn” for more information.
22
471
If you find problems with this program, report them with trouble(1L).
This particular message comes from /usr/local/lib/rn/newsnews.
will only see it once.
You
[Type space to continue] _
Because I haven’t used the program before, this first time out will have all sorts of
information:
Trying to set up a .newsrc file--running newsetup...
Creating /users/taylor/.newsrc to be used by news programs.
Done.
If you have never used the news system before, you may find the articles in
news.announce.newusers to be helpful. There is also a manual entry for rn.
To get rid of newsgroups you aren’t interested in, use the u command.
Type h for help at any time while running rn.
(Revising soft pointers--be patient.)
Unread news in general
Unread news in news.admin.misc
Unread news in news.admin.policy
Unread news in news.admin.technical
Unread news in news.announce.conferences
etc.
etc.
1
88
68
6
26
article
articles
articles
articles
articles
Finding new newsgroups:
********
CAUTION
1 unread article in general--read now? [ynq] _
One of the worst aspects of rn is that, by default, it subscribes you to over
2,400 newsgroups the first time you enter the program. Fortunately, the fix
is easy.
3. Now that I’m a member of a few thousand groups, I want to get out of some! The
fastest way for me to fix the problem is to quit rn (type q), then use vi to edit my
personal Usenet database file, called .newsrc, which resides in my home directory.
This file contains a list of all newsgroups, with each group followed by a special
character and an indication of which articles I have already seen:
% head .newsrc
general: 1-699
news.admin: 1-26982
news.admin.misc: 1-6329
news.admin.policy: 1-8171
news.admin.technical: 1-445
22
Hour 22
472
news.announce.conferences: 1-5326
news.announce.important:
news.announce.newgroups:
news.announce.newusers:
news.answers:
To unsubscribe and have no groups included, I simply replace the colon on each
line with an exclamation point. A group that isn’t subscribed looks like this:
news.admin! 1-26982 . You can use :1,$s/:/!/ in vi to unsubscribe quickly to all
groups. I suggest, however, that you make sure that you do read the group general
on your system.
4. Now, I start rn a second time, and here’s what I see:
% rn
********
1 article
Instead of answering yes or no to this question, I can go to the groups I’d previously chosen to read, using the g groupname command:
1 unread article in general--read now? [ynq] g comp.sys.mac.announce
Newsgroup comp.sys.mac.announce is currently unsubscribed to-➥resubscribe? [yn]
It’s not surprising that I am not subscribed: I just unsubscribed from almost
everything. Joining the group sounds good, so I type y, and the screen changes:
0 unread articles in comp.sys.mac.announce--read now? [ynq] _
There are no articles pending in this newsgroup, as can be seen by the 0 unread
message. That’s good news—I then can add the other three groups by saying n
here, because I do not want to read this group, then using the g command with
each of the other groups specified.
The first time I choose n, however, the prompt changes to a new line:
End of newsgroups--what next? [qnp] _
I can enter the g newsgroup here just as easily.
5. Imagine it’s a couple days later when I log in to my computer and again enter rn to
see if there’s any news. Unsurprisingly, quite a few articles have arrived since I
signed up for the groups:
% rn
Unread news in sci.edu
Unread news in misc.writing
1 article
28 articles
81 articles
I start by saying n, then y when it asks about the newsgroup sci.edu. The screen
changes:
22
Article 4695 (27 more) in sci.edu:
From: [email protected] (Henry Choy)
Subject: Re: Automation and the Need for Education
Date: 9 Dec 1993 22:00:10 GMT
Organization: University of Saskatchewan
NNTP-Posting-Host: sparkle.usask.ca
X-Newsreader: TIN [version 1.2 PL0]
Alberto Moreira ([email protected]) wrote:
:
Computers can’t do anything by themselves - they are raw
:
metal. But the computer program running in a computer is
:
a transliteration of a human creator’s thought processes.
:
When you play chess against a computer, you’re not playing
:
against the computer; you’re playing against the individuals
:
who designed Sargon, Deep Thought, or whatever chess program
:
you’re battling.
In a way this is so, but the designers of these programs are working in
the way they were taught to play chess :)
Consider a chess program that is free to consider different evaluation
techniques, and other whatnot procedures and measures concocted with
some element of randomness. This program behaves
--MORE--(46%) _
There’s a lot of information contained in the first few lines, as shown in Figure
22.1. Also, the lines that begin with a colon are quoted text from a previous article
in the group, the one written by Alberto Moreira. All lines not beginning with a
colon are the thoughts of Henry Choy, the author of this particular article.
Figure 22.1.
Information on rn article
display screen.
473
22
Hour 22
474
6. While reading a particular article, I can choose from a variety of different commands that I can enter to perform different actions, most notably q to quit reading
this particular article, n to move to the next article, k to mark this article and all of a
similar subject as read (so that I don’t see them), ^S (or space) to read the next
article with the same subject in this group, R to reply to the author via electronic
mail, and F to follow up this article with thoughts and reactions of my own. One
particularly helpful command is =, which offers a table of contents for this group,
showing all articles I have not yet read.
TIME SAVER
One little-known fact about rn is that you can define exactly what is
shown in the = table of contents screen with the SUBJLINE environment
variable. I have mine set to %t -- %s, which produces a list with both the
author and subject indicated.
When I enter =, here’s what happens:
4696 [email protected] -- INTERACTIVE LEARNING (INTERACT 2)
4697 [email protected] -- Academic tenure in Australia and
other countries
4698 [email protected] -- Grade Point Averages: What’s the point?
4699 [email protected] -- Grade Point Averages: What’s the
point?
4700 [email protected] -- Grade Point Averages: What’s
the point?
4701 [email protected] -- Looking for U of A - John Lind...
point?
other countries4704 [email protected] -- TOSHIBA/NSTA contest info
other countries
4709 [email protected] -- Q: TV use in education!
point?
4711 [email protected] -- Subscription to Russian literature magazine in
English and in Russian!
4712 [email protected] -- PhD & MA Programs in Learning Sciences at
Northwestern
4713 [email protected] -- Help needed on paper
[Type space to continue] _
There are more articles waiting to be read than can fit on the screen (hence the
Type space to continue at the bottom). Here, I can type q to zip back to the
22
475
bottom of the article I was reading (then M to mail a response to the author,
perhaps) or enter a specific article number to read that article. Article 4704, titled
TOSHIBA/NSTA contest info, sounds pretty interesting, so I’ll enter 4704 and move
directly to that article:
Article 4704 (23 more) in sci.edu:
Subject: TOSHIBA/NSTA contest info
Date: 13 DEC 93 02:31:15 CST
Organization: University of Winnipeg
NNTP-Posting-Host: uwpg02.uwinnipeg.ca
Hi. I was wondering if anyone had sample material
regarding the NSTA/TOSHIBA explorovision contest.
SPecifically, sample storyboards and written reports.
Thanks for any info. M. Carroll U of Wpg
internet: [email protected]
End of article 4704 (of 4719)--what next? [^Nnpq]_
At this point, again, I could use R to reply, F to post a follow-on article of my own,
or any of the other commands. It seems like a good time to quit Usenet for now, so
I type q and move to the next group in the list:
End of article 4704 (of 4719)--what next? [^Nnpq]
********
81 unread articles in misc.writing--read now? [ynq] _
Another q, and I’m back at my C shell prompt.
7. Just for comparison, here’s what tin looks like when I launch it on the list of
newsgroups that I actually read on an approximately daily basis:
Group Selection (18)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
396
16
2
38
447
158
110
603
37
20
h=help
misc.forsale.computers.mac
Apple Macintosh related comput
news.announce.important
General announcements of inter
news.announce.newusers
Explanatory postings for new u
comp.binaries.mac
Encoded Macintosh programs in
comp.org.sug
Talk about/for the The Sun Use
comp.org.usenix
USENIX Association events and
comp.sys.mac.announce
Important notices for Macintos
comp.sys.sun.announce
Sun announcements and Sunergy
rec.arts.movies.reviews
Reviews of movies. (Moderated)
rec.food.recipes
Recipes for interesting food a
alt.books.reviews
“If you want to know how it tu
alt.education.distance
Learning over nets etc.
alt.education.research
alt.folklore.herbs
misc.education
Discussion of the educational
misc.education.language.english
22
Hour 22
476
<n>=set current to n, TAB=next unread, /=search pattern, c)atchup,
g)oto, j=line down, k=line up, h)elp, m)ove, q)uit, r=toggle all/unread,
s)ubscribe, S)ub pattern, u)nsubscribe, U)nsub pattern, y)ank in/out
_
Many of these groups have no articles pending, but you can see (on the second
number on the line) that the group misc.forsale.computers.mac has 396 new
articles, comp.org.sug has 16, and so on. To read the currently highlighted group, I
press Return.
misc.forsale.computers.mac (318T 391A 0K 0H R)
**
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
1
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+ 2
4
2
2
Group Purchase: SCSI Hard Drives
BernulliMD150+5(150)carts!$750NEW!
MUST SELL: Macintosh external AppleCD 150
Mac Centris 650 Forsale
IBM/Tandberg TDC3600/250Mb SCSI Tape Drives
FOR SALE: Symantec C++ 6.0.1
SCSI stuff for sale
Computer Accessaries for Sale
DOS 6.0 FOR SALE
Performa 600 (Max IIVx NOT!)
WANTED: color ‘040 mac used.
Ashlar Vellum 2D and 3D for sale
Cheap ImageWriter II for sale $95 (shipping
PB 140
Apple CD150 CD_ROM, new $199.00
LCII 8/80 - Gotta sell soon.
h=help
David La Croix
Donn Lasher
Seung Jong Lee
M Torricelli
[email protected]
[email protected]
Victor Mark Kalyuz
Qian Zhang
Qian Zhang
Steve Fouts
[email protected]
robert dornbusch
Prachya Chalermwat
Jeffrey Ello
Victor Mark Kalyuz
Anthony S. Kim
**
<n>=set current to n, TAB=next unread, /=search pattern, ^K)ill/select,
a)uthor search, c)atchup, j=line down, k=line up, K=mark read, l)ist thread,
|=pipe, m)ail, o=print, q)uit, r=toggle all/unread, s)ave, t)ag, w=post
_
I press Enter again, and I can read the specific article, or use q to quit. I think I’ve
had enough Usenet for one day, so I quit and again return to my C shell prompt.
Usenet is a complex, wild, and almost organization-free set of thousands of newsgroups
distributed to over 100,000 systems and read by millions of users throughout the
world. Between Mac, PC, Amiga, UNIX, and VMS, there are at least 20 different programs
available just for reading the vast volumes of information that flow past each day. If you have
access to a Mac or a PC, or even a UNIX system running the X Window System, there are
some attractive programs available for reading the Net. If not, I’m a fan of tin and also have
22
used rn for many years; both will serve you well once you learn them. Certainly, for any of
these programs, start out with the included documentation, or at least read through the
(usually quite long) man pages.
Gopher and e-mail are even more wild, but they’re all very useful pieces of the Internet and
UNIX pie, and, if I had to point to a single program as the “killer application” of the Net,
it’d be e-mail. My advice, after almost 17 years of being involved with the UNIX and Internet
communities, is that you can’t go wrong by becoming extremely comfortable with your
favorite e-mail program, and ditto for Usenet.
Workshop
in the next hour.
Key Terms
domain naming Domain naming is the addressing scheme for hosts on the Internet. The
domain name is the information after the hostname on the right side of the @ in an address.
For example, [email protected] has a domain name cornell.edu and a full domain name
mutt.cornell.edu .
gopherspace The information space through which gopher travels while you’re using the
program.
newsgroup A Usenet group focused on a particular subject or topic of conversation.
protocol An agreed-upon language for transfer of information between two computers.
quoted text A portion of a previous article that is included in the current article to give
context, particularly in disagreeing with or amplifying specific thoughts.
Questions
1. Use finger to see whether there’s anyone logged in at one of the systems in the list,
and then use it to find out more information about a person on that list of users.
2. If you have a friend on your system or another system, use talk to say hello.
3. Using archie, find one or two archive sites that have a copy of the tin newsreading
program.
4. What does the acronym LISP stand for? Use the WAIS system to find out.
5. Check in to the library at Stanford University with gopher and see whether it has a
copy of this book yet.
477
22
478
Hour 22
6. Enter the findgroup alias and find the groups that discuss the following:
plant biology
laser printers
UNIX questions
anthropology
7. Using rn or another news reader, check in to one of the groups you just found, and
see whether there are any new articles. Based on information in this hour, how
many systems do you think get this group?
In the next hour, you continue with the tour of the Internet. You learn how to use archie,
ftp, and telnet.
22
Hour
23
In this hour, you pick up where you left off, with a tour of Internet facilities.
Goals for This Hour
■
■
■
■
Connecting to remote Internet sites
Copying files from other Internet sites
Finding archives with archie
A few interesting telnet sites
This hour is intended to offer a quick and enjoyable overview of the many
services available through the Internet. From finding that long-lost archive to
exploring hotels available overseas, the range of information available will
undoubtedly astound you!
479
23
Hour 23
480
Task 23.1: Connecting to Remote Internet Sites
The really fun part of UNIX, and one reason that it’s grown dramatically in the last
few years, is that it’s the most connected operating system in the world. The variety
of different services available for users of a networked UNIX machine is staggering. Not only
can you use all the commands explained in this hour, but you can use some additional
services, such as gopher and archie, that are on the cutting edge of information services and
that will be explained in the final hour.
At its simplest, the connectivity all relies on very high-speed wires coming out of the back of
the computer you’re on and connecting to other computers. Unlike the telephone-based
connections of UUCP, this line is always alive and is much, much faster, able to stream
literally megabytes of information in under a minute. The big network itself, the Internet,
evolved from an earlier network called the ARPAnet, funded by the Advanced Research
Projects Agency of the U.S. Government in the 1970s and into the 1980s. Somewhere along
the way, it began to grow beyond the vision and capabilities of the original design and began
being known as the ARPA Internet. In the last few years, the government (and particularly
the National Science Foundation) has begun to withdraw from its overseer role, and as a
result, the system is now known as the Internet.
If you’ve heard of the Information Highway (which is now being called the Information
Superhighway), you’ve also heard about the Internet, which is the existing roadway that is
growing to become this high-speed thoroughfare of information. In fact, next time a friend
mentions the Information Highway, you can say that you’ve already experienced some of it
by working on the Internet.
There are three main tasks that the Internet can help you with: using remote systems, sending
mail to remote users, and working with remote file systems. In addition, you can find out who
is logged on to any system on the Internet and use the talk program to talk with someone else.
If you know that the remote site is a UNIX system, the easiest way to log in to that site is to
use the rlogin command, which has the awkward notation of rlogin host -l account. If
you aren’t sure about the system, use telnet, which is the universal program for connecting
to remote computer systems. Unlike any of the other programs you’ve learned so far, telnet
actually works either as a simple program you can invoke from the command line, or as a
sophisticated environment for connecting to various systems.
1. First off, I’ll use rlogin to connect to a remote system and see if I have a file there:
% rlogin netcom.com
Password:_
23
481
By default, rlogin assumes that your account on the remote system has the same
name as your account on your home system. If you forget to use the -l account
option, just press Return here, and it prompts for an account name:
% rlogin netcom.com
Password:
Login incorrect
login: taylor
Password:_
Once I enter my password, I’m logged in to the remote system:
Last login: Mon Dec 13 09:38:35 from utech
SunOS Release 4.1.3 (NETCOM) #1: Wed Sep 23 05:06:55 PDT 1996
NETCOM On-line Communication Services, Inc.
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
Washington DC: Additional modems have been added.
Santa Cruz: Additional modems have been added.
Elm has been updated to version 2.4 (PL 21).
using elm with aliases should run the
“/usr/local/bin/newalias” program.
Users
%
Using ls tells me what I want to know:
netcom % ls
Global.Software
Interactive.Unix
Mail/
netcom %
News/
Src/
bin/
history.usenet.Z
login
testme
2. The rlogin command offers a shorthand notation for logging out of the remote
system; instead of logout, you can simply enter ~.. To stop the rlogin session, use
~^z. No other tilde commands are available in rlogin .
I choose to log out the normal way:
netcom % logout
Connection closed
%
Now I’m back on the original computer system.
3. The alternate way to connect to a remote computer is to use telnet. The easiest
way to use this command is the same way you use rlogin. At the command
prompt, specify the name of the system to which you want to connect:
% telnet netcom.com
Trying...
Connected to netcom.com.
23
Hour 23
482
SunOS UNIX (netcom)
login: _
Notice that this way is much more like having a terminal connected to this system.
I can log in, enter my password, and then have a new login session on the remote
system as if I were sitting in that computer room working away.
4. Instead, though, I’m going to use the ^] control character to switch into the telnet
program itself:
SunOS UNIX (netcom)
login: ^]
telnet > _
Now I enter help to see what the options are:
telnet> help
Commands may be abbreviated.
close
display
mode
open
quit
send
set
status
toggle
z
?
telnet>
Commands are:
close current connection
display operating parameters
try to enter line-by-line or character-at-a-time mode
connect to a site
exit telnet
transmit special characters (‘send ?’ for more)
set operating parameters (‘set ?’ for more)
print status information
toggle operating parameters (‘toggle ?’ for more)
suspend telnet
print help information
_
There are lots of possible commands. I choose to return to my connection to
Netcom, however, so I just press Return, and I’m back at the login prompt. If I
don’t enter anything quickly enough, the remote system automatically drops the
connection:
login: Login timed out after 60 seconds
Connection closed by foreign host.
%
To log out of the remote system, the best strategy is simply to exit the telnet
session, which will drop the line automatically. If that doesn’t work, the ^]
sequence followed by either quit or close will do the trick.
5. To start out directly in the telnet command mode, simply enter the command
without specifying a remote host:
% telnet
telnet> _
From here, connecting to the remote host is also quite simple:
telnet> open netcom.com
Trying...
23
483
SunOS UNIX (netcom)
login: _
Again, I use ^] and close to close the connection.
Both the rlogin and telnet commands are useful in different situations, but I find
myself using the rlogin command more often because it sends much of the current
environment along to the remote system. So if I have my system set for a specific type of
terminal (that is, the TERM variable is set to a specific value), that value is automatically copied
into the new environment of the remote system, which saves lots of hassle.
Task 23.2: Copying Files from Other Internet Sites
The main program used to copy files on the Internet is ftp, which is named after the
protocol it implements, the file transfer protocol. Like much of UNIX, ftp can take
a while to master, particularly because no effort has been made to make it at all user-friendly.
Nonetheless, it functions very similarly to the telnet command; you either enter ftp to start
the program and then specify the system with which you’d like to connect, or you specify the
name of the system on the command line. Either way, you are then prompted for an account
and password; then you are dropped into the ftp prompt with the connection open and
waiting.
Many sites talk about having anonymous ftp capabilities. Systems allowing this connection
indicate that you don’t need your own computer account on that machine to be able to
connect and copy files from their archives. To use these systems, enter ftp as the account
name, and then enter your own e-mail address as the password (that is, I’d enter ftp and then
[email protected] as the password). The most important commands available in ftp are
summarized in Table 23.1. The most important one to remember is bye, which you use when
you’re done.
Table 23.1. Valuable ftp commands.
Command
Meaning
ascii
Set ftp to transfer a text (ASCII) file.
Set ftp to transfer a binary file, probably a program or database of
information.
Quit the ftp program.
Change the remote directory to dir.
Close the current connection.
binary
bye
cd dir
close
continues
23
Hour 23
484
Command
dir
get
lcd dir
ls
mget
mput
open
prompt
put
pwd
Meaning
Print a listing of files in the current remote directory.
Transfer a file from the remote system to your local system.
Change the current directory on the local system to dir or to your
home directory if no argument is given.
List the files in the current remote directory.
Multiple get—get files with a wildcard matching capability.
Multiple put—put files with a wildcard matching capability.
Open a connection to the specified remote machine.
Control whether or not to ask for confirmation of each file transferred if using mget or mput.
Put a file onto the remote system from the local system.
Show the present working directory on the remote.
1. To start out, I want to pick up a file from netcom that I saw earlier when I used
rlogin to look at the remote system. To start ftp, I use the short notation of
specifying the host at the command line:
% ftp netcom.com
220 netcom FTP server (Version 2.1 Fri Apr 9 13:43 PDT 1996) ready.
Name (netcom.com:taylor): _
By default, ftp assumes that I want to use the same account name, which in this
case I do, so I press Return, and then enter my password:
Name (netcom.com:taylor):
331 Password required for taylor.
Password:
230 User taylor logged in.
ftp> _
2. Now I’m at the ftp program prompt, and any of the commands shown in Table
23.1 will work here. To start, I use dir and ls to list my files in different formats:
23
ftp> dir
200 PORT command successful.
150 Opening ASCII mode data connection for /bin/ls.
total 140
-rwxr-xr-x 1 taylor
users0
4941 Oct 4 1991
-rwx------ 1 taylor
daemon
987 Sep 20 1992
-rw-r--r-- 1 taylor
users0
2103 Sep 30 19:17
-rw-r--r-- 1 taylor
users0
752 Apr 17 1992
-rw-r--r-- 1 taylor
users0
1749 Jun 8 1993
.Pnews.header
.accinfo
.article
.cshrc
.delgroups
drwx------ 2 taylor
daemon
4096 Dec 6 14:25
-rw-r--r-- 1 taylor
users0
28 Nov 5 09:50
-rw------- 1 taylor
users0
0 Jun 9 1993
-rw-r--r-- 1 taylor
users0
1237 Dec 13 09:40
-rw-r--r-- 1 taylor
users0
6 Aug 6 1991
-rw-r--r-- 1 taylor
users0
538 Dec 6 14:32
-rw-r--r-- 1 taylor
users0
537 Dec 6 14:30
-rw-r--r-- 1 taylor
users0
1610 Feb 17 1992
-rw-r--r-- 1 taylor
users0
0 Aug 6 1991
-rw-r--r-- 1 taylor
users0
45 Feb 2 1993
-rw-r--r-- 1 taylor
users0
6 Feb 8 1993
-rw-r--r-- 1 taylor
users0
16767 Jan 27 1993
-rw-r--r-- 1 taylor
users0
114 Apr 6 1992
drwxr-xr-x 4 taylor
users0
4096 Nov 13 11:09
-rw-r--r-- 1 taylor
users0
1861 Jun 2 1992
-rw------- 1 taylor
users0
22194 Oct 1 1992
drwx------ 4 taylor
users0
4096 Nov 13 11:09
drwxr-xr-x 2 taylor
users0
4096 Nov 13 11:09
drwxr-xr-x 2 taylor
users0
4096 Nov 13 11:09
drwxr-xr-x 2 taylor
users0
4096 Nov 13 11:09
-rw-r--r-- 1 taylor
users0
12445 Sep 17 14:56
-rw-r--r-- 1 taylor
users0
1237 Oct 18 20:55
-rw-r--r-- 1 taylor
users0
174 Nov 20 19:21
226 Transfer complete.
1792 bytes received in 3.1 seconds (0.56 Kbytes/s)
ftp> ls
150 Opening ASCII mode data connection for file list.
.cshrc
.login
.elm
Mail
News
.logout
.newsrc
.rnlast
.rnsoft
bin
.tin
Global.Software
.sig
.oldnewsrc
.pnewsexpert
.plan
.Pnews.header
history.usenet.Z
.rnlock
Src
.ircmotd
.article
.delgroups
.accinfo
.forward
Interactive.Unix
testme
login
269 bytes received in 0.02 seconds (13 Kbytes/s)
ftp>
485
.elm
.forward
.ircmotd
.login
.logout
.newsrc
.oldnewsrc
.plan
.pnewsexpert
.rnlast
.rnlock
.rnsoft
.sig
.tin
Global.Software
Interactive.Unix
Mail
News
Src
bin
history.usenet.Z
login
testme
23
Hour 23
486
As you can see, ftp can be long-winded.
JUST A MINUTE
One trick for using the ls command within ftp is that if you specify a set
of command flags as a second word, it works fine. Specify a third
argument, however, and it saves the output of the command into a local
file by that name; so ls -l -C would create a file called -C on your
system with the output of the ls -l command.
Because you can supply some flags to the ls command, I always use -CF to force
the output to list in multiple columns and show directories, which makes the
output readable:
ftp> ls -CF
.Pnews.header*
.newsrc
Interactive.Unix
.accinfo*
.oldnewsrc
Mail/
.article
.plan
News/
.cshrc
.pnewsexpert
Src/
.delgroups
.rnlast
bin/
.elm/
.rnlock
history.usenet.Z
.forward
.rnsoft
login
.ircmotd
.sig
testme
.login
.tin/
.logout
Global.Software
remote: -CF
ftp>
3. To transfer the file login from the remote system, I can use the get command:
ftp> get
(remote-file) login
(local-file) login.netcom
150 Opening ASCII mode data connection for login (1237 bytes).
local: login.netcom remote: login
ftp>
4. Alternatively, I could use mget and specify a wildcard pattern similar to one I’d give
the shell:
ftp> mget log*
mget login? y
150 Opening ASCII mode data connection for login (1237 bytes).
local: login remote: login
ftp>
23
487
There was only one match, so the transfer was easy. Entering anything other than y
at the mget login? prompt would have resulted in the file not being transferred.
That was easily accomplished. Now I will look on another system in the anonymous FTP directory to see what’s available.
5. To disconnect, I enter close so that I don’t leave the ftp program:
ftp> close
221 Goodbye
ftp>
There are hundreds of information servers on the Internet, offering an astounding
variety of information, from weather service maps to the full text of The Bible and
Alice in Wonderland to the source listings of thousands of different programs.
In this example, I want to look at the anonymous FTP archive at the Massachusetts
Institute of Technology’s Artificial Intelligence Laboratory. The host is called
ftp.ai.mit.edu :
ftp> open ftp.ai.mit.edu
Connected to mini-wheats.ai.mit.edu.
220 mini-wheats FTP server (Version 2.1b Wed Aug 25 09:20 EDT 1993)
Name (ftp.ai.mit.edu:taylor): ftp
331 Guest login ok, send your complete e-mail address as password.
Password:
230230230-Welcome to the MIT Artificial Intelligence Laboratory. If you are
230-interested in Artificial Intelligence Laboratory publications please
230-ftp to publications.ai.mit.edu.
230230230230 Guest login ok, access restrictions apply.
ftp>
Now I can use ls
-CF
to look around:
ftp> ls -CF
.message
bin/
etc/
pub/
ai-pubs/
dev/
incoming/
usr/
remote: -CF
ftp>
It looks like there might be something of interest in the pub directory (a directory
by this name usually contains public information). I use cd to change to that
directory, then ls -CF to see what’s available there:
ftp> cd pub
250 CWD command successful.
ftp> ls -CF
23
488
Hour 23
6.824/
medical-white-paper.ps.Z*
Address:
memtr
BL.tar.Z
minsky/
GA/
misc/
ICv2.45.sit.hqx
mit1345.tar
Iterate/
mobile-dist-telecomp/
MC132p_structures.cif
mobot-survey.text
MSV_array.cif
mr-sd-mapped.ps
MSV_structures.cif
mr-sd.ps
Peng_Wu_Thesis.ps.Z
mtm/
README
ontic/
TS/
patches.c
aal
pdp8-lovers-archive
adage/
pgs-th.ps.Z
ai3/
pinouts/
aimr/
poker/
akcl.Z
psabalone.tar.Z
alan@
pset32new.c
aop/
publications/
ariel/
qobi/
autoclass/
ra.ps
bson/
rbl-94.archive
cki/
refer-to-bibtex/
clmath.tar
sanger-figures/
cube-lovers/
sanger-papers/
cva/
sanger.mackey.tar.Z
cwitty/
sanger.mackey.tar.gz
dam/
scheme-libraries/
doc/
screamer/
dssa/
screamer.tar.Z
dssa.ps
screamer3.04/
eel/
screen/
ellens@
series/
engine/
square-dancing/
fax/
squash-ladder
hebrew/
surf-hippo/
incoming@
swill
iter-man.ps
swillcoxswillcoxswillcoxAddress:
iterate.lisp
systems/
iterate.tar.Z
tandems@
jupiter/
tbs/
lemacs/
texture.tiff
linalg.shar
tgif.tar.Z
lisp3/
transition-space.lisp
logtalk.uue.Z
turing_option
loop-macro.tar
turing_option.ps
lptrs/
users/
ltalk*
viola-fbr.ps.Z
lyskom-0.33.1.english.el.Z
vis/
maddog/
who-line-gc-thermometer.lisp@
medical-white-paper.dvi
x3j13/
remote: -CF
ftp>
23
489
6. A README file is usually a good thing to start with. A handy ftp trick is that you can
copy files directly to your screen by using /dev/tty as the local filename, or you can
even pipe them to programs by using the pipe symbol as the first character:
ftp> get README |more
150 Opening ASCII mode data connection for README (186 bytes).
This file will be expanded eventually.
If you’ve been looking at directories of specific users, such as ‘ian’
or ‘ellens’, those directories have been moved into the users
directory.
local: |more remote: README
ftp>
In this case, the README file is not incredibly helpful.
7. It’s time to split and check another FTP archive, this time one at Apple Computer
(ftp.apple.com):
ftp> close
221 Goodbye.
ftp> open ftp.apple.com
Connected to bric-a-brac.apple.com.
220 bric-a-brac.apple.com FTP server (IG Version 5.93 (from BU,
from UUNET 5.51) Sun Nov 21 14:24:29 PST 1993) ready.
Name (ftp.apple.com:taylor): ftp
331 Guest login ok, send ident as password.
Password:
230 Guest login ok, access restrictions apply.
ftp>
Again, ls
-CF
shows what files are available:
ftp> ls -CF
.cshrc
alug/
boot/
echt90/
.login
apda/
cdrom/
etc/
.logout
apple/
dev/
pie/
README
bin/
dts/
pub/
remote: -CF
ftp>
Use the |more trick to see what the README file has to say:
ftp> get README |more
150 Opening ASCII mode data connection for README (424 bytes).
This is the top level of our FTP server.
public/
shlib/
software/
23
Hour 23
490
If you’re an authorized person, and you want to have a directory on the
top level, please contact Erik Fair <[email protected]>, (408) 974-1779,
and explain why you want one, and what you’re going to use it for.
Otherwise, please make your stuff available in the “public” directory.
There are no writeable directories for anonymous FTP on this server; it
cannot be used as a drop box.
local: |more remote: README
ftp>
8. There’s a new Macintosh application available on this system in dts/mac/hacks that
I’ve been interested in seeing. I can move directly there with cd, confirming that
I’m where I think I am with pwd:
ftp> cd dts/mac/hacks
250 CWD command successful.
ftp> pwd
257 “/dts/mac/hacks” is current directory.
ftp> ls -CF
aetracker-3-0.hqx*
mountalias-1-0.hqx*
applicon-2-1.hqx*
newswatcher.hqx*
appmenu-3-5.hqx*
okey-dokey-1-0-1.hqx*
bison-flex.hqx*
oscar.hqx*
colorfinder.hqx*
piston.hqx*
darkside-of-the-mac-4-1.hqx*
snake.hqx*
dropper.hqx*
switchapp-1-1.hqx*
escape-dammit-0-4.hqx*
system-picker-1-0.hqx*
extensions-manager-2-0-1.hqx*
thread-manager-exten-1-2.hqx*
flipper.hqx*
trashman-4-0-2.hqx*
folder-icon-maker-1-1.hqx*
understudy.hqx*
fsid.hqx*
unlockfolder.hqx*
im-mac-1-0b26w.hqx*
virtual-controllers.hqx*
lockdisk-1-0.hqx*
xferit-1-4.hqx*
remote: -CF
ftp>
Notice that there’s an asterisk following the names of these files. This indicates, as
you know from the -F flag to ls, that it’s a binary file. Therefore, I need to specify
to ftp that it should transfer the file in binary mode, by entering binary:
ftp> binary
200 Type set to I.
ftp>
I check to see how big the file is, then I can use get to transfer it and drop the
connection with bye:
ftp> dir colorfinder.hqx
23
491
-rw-r-xr-x 1 mjohnson archivis
43442 May 24 1991 colorfinder.hqx
remote: colorfinder.hqx
71 bytes received in 0 seconds (0.069 Kbytes/s)
ftp> get colorfinder.hqx
150 Opening BINARY data connection for colorfinder.hqx (43442 bytes).
local: colorfinder.hqx remote: colorfinder.hqx
ftp> bye
221 CUL8R.
%
23
Now that I’m back at the command prompt, I can use ls again to confirm that I’ve
received both the colorfinder.hqx and login.netcom files:
% ls
Archives/
InfoWorld/
Mail/
News/
OWL/
awkscript
%
bin/
buckaroo
buckaroo.confused
cheryl
colorfinder.hqx
dickens.note
keylime.pie
login.netcom
newsample
papert.article
sample
sample2
sample3
src/
src.listing
temp/
tmp.listing
who.is.who
The FTP system is a terrific way to obtain information from the Internet. Thousands
of systems offer various services via anonymous FTP, too. Table 23.2 lists a few of the
most interesting ones.
Table 23.2. Some interesting FTP archives.
Site
Institution and Available Information
aenas.mit.edu
Massachusetts Institute of Technology Free Software
Foundation site. Files: GNU EMACS
University of Georgia. Files: LISP, PROLOG, natural
language processing, MS-DOS utilities
University of Nevada. Files: U.S. Constitution and
supporting documents, religious texts, the Bible
Brown University. Files: Mac
San Francisco University. Files: MS-DOS, Mac
Michigan State University. Files: MS Windows
Rice University. Files: Sun-Spots, Amiga, ispell, ofiles
California State University, Chico. Files: online
chemistry manual
U.S. National Institute of Health
aisun1.ai.uga.edu
archive.nevada.edu
brownvm.brown.edu
cc.sfu.ca
clvax1.cl.msu.edu
cs.rice.edu
cscihp.ecst.csuchico.edu
cu.nih.gov
continues
Hour 23
492
Site
Institution and Available Information
deja-vu.aiss.uiuc.edu
University of Illinois Champaign-Urbana. Files: Rush
Limbaugh transcripts, Monty Python, humor, song
lyrics, movie scripts, urban legends
University of Kentucky. Files: Mac, MS-DOS, UNIX,
Amiga, NeXT, 386BSD, AppleII, GNU, RFCs, various
Usenet archives
Apple Computer. Files: Apple (Mac, II, IIgs) product
information, software, developer support
Indiana University. Files: UNIX, MS-DOS, NeXT
updates, MS Windows 3.x archive
Liverpool University Computer Science Department.
Files: Ports to HP-UX machines (especially Series 700),
including X11R4 clients, GNU, recreational software,
text editors, system administrator tools
Electronic Frontier Foundation.
Digital Equipment Corporation, Palo Alto, California.
Files: X11, recipes, cron, map, Modula-3
California Institute of Technology. Files: GNU (Free
Software Foundation)
Hewlett-Packard, Corvallis, Oregon. Files: Motif, archives
University of Maryland. Files: government-related,
books, economics, MS-DOS, Novell, Mac
University of California, Santa Cruz. Files: amoeba, U.S.
Constitution
National Science Foundation Network. Files: Network
information, Internet Resource Guide
NASA. Files: Hubble space telescope images
San Diego State University. Files: sounds
SRI International Files: improving the security of your
UNIX system
Cornell University. Files: tcsh
FTP software. Files: FTP-related programs
Columbia University. Files: kermit
U.S. Army—White Sands Missile Range. Files:
MS-DOS, UNIX, CPM, Mac
f.ms.uky.edu
ftp.apple.com
ftp.cica.indiana.edu
ftp.csc.liv.ac.uk
ftp.eff.org
gatekeeper.dec.com
hobiecat.cs.caltech.edu
hpcvaaz.cv.hp.com
info.umd.edu
midgard.ucsc.edu
nnsc.nsf.net
nssdca.gsfc.nasa.gov
sciences.sdsu.edu
sparkyfs.erg.sri.com
tesla.ee.cornell.edu
vax.ftp.com
watsun.cc.columbia.edu
wsmr-simtel20.army.mil
23
493
There’s no question that the interface to ftp is awkward, however, and there are a couple of
different programs that have tried to address this problem, as you learn in the last hour in this
book.
There’s no way here to fully cover all the information available on the Internet, so if you’re
excited by these possibilities, I strongly recommend that you obtain a copy of the book
Navigating the Internet by Mark Gibbs and Richard Smith. It’s a terrific introduction to the
many services available on the Internet, including gopher, archie, the World Wide Web,
telnet, and ftp. Those that aren’t covered in this hour are shown in the last hour of this book,
however; so you at least will have seen most of these important Internet information-related
commands.
Task 23.3: Finding Archives with archie
If you spent any time at all looking at the list of FTP archives in the previous task,
you already realize that just obtaining a list of files available in one system can be quite
a chore. Yet, computers are ideally suited to serve as their own navigational aides, as they easily
work with large and complex databases.
The archie system was developed at McGill University in Canada. It is a huge database of
all files and directories available on all registered FTP sites in the world. That’s quite a bit:
Over 2.5 million different files are in the database!
Nonetheless, archie is a fairly simple-minded program, and there’s only so much information you can glean by being able to analyze just file and directory names. For example, if I
have a program called “Wallpaper Demo for the Mac” and save it in a file wallpaper.demo.MAC,
odds are pretty good that people who search for Macintosh demonstration programs could
find it. What if I decided that was too many letters and instead named it wp.mac? It’s much
less likely that folks would know what the file contains.
There are a few options worth knowing before using archie, most notably that the format
of the program itself is archie search-string. By default, the program lists only exact
matches to the pattern, but -c forces it to match on either upper- or lowercase letters,
depending on the pattern; -e forces exact matches (this is the system default, but some sites
have other default actions—it’s up to your local system administrator); -s considers the
search pattern as a possible substring; -r searches for the specified regular expressions; and
-l lists the results in a format suitable for use with other programs (such as fget). One final
option is of interest: The -L option lists all archie servers known to the program.
CAUTION
If you don’t have the archie program on your system, don’t despair! You
can use telnet to connect to archie.rutgers.edu, archie.sura.net, or
archie.unl.edu to interact with the archie databases directly (although
23
Hour 23
494
this isn’t necessarily faster than using the archie program if you have it on
your system!). Finally, if you aren’t on the Internet at all, you can send
electronic mail to archie at any of the three systems listed. Use prog
search-string , and ensure that the last line of your message is quit so it
knows when to stop reading your mail for commands.
1. To start out, I want to see what archie servers are known by my version of archie:
% archie -L
Known archie servers:
archie.ans.net (USA [NY])
archie.rutgers.edu (USA [NJ])
archie.sura.net (USA [MD])
archie.unl.edu (USA [NE])
archie.mcgill.ca (Canada)
archie.funet.fi (Finland/Mainland Europe)
archie.au (Australia)
archie.doc.ic.ac.uk (Great Britain/Ireland)
archie.wide.ad.jp (Japan)
archie.ncu.edu.tw (Taiwan)
* archie.sura.net is the default Archie server.
* For the most up-to-date list, write to an Archie server and give it
the command `servers’.
Notice the third line from the end: My default archie server is archie.sura.net.
The other servers listed can be accessed but aren’t checked directly. Usually, it
doesn’t matter which server is used because the information available through
different servers is mostly identical. If you just know something’s out there but
can’t find it, check a few different servers with archie -h servername.
2. To search for a specific program, I simply can enter the name of the program. I’m
interested in finding a UNIX program called newmail:
% archie newmail
Host plaza.aarnet.edu.au
Location: /usenet/comp.sources.unix/volume25
FILE -r--r--r-15049 Dec 20 1991
newmail
Host gum.isi.edu
Location: /share/pub/vmh/bin
FILE -rwxr-xr-x
104
Jul
9 18:26
newmail
104
Jul
9 11:26
newmail
Host venera.isi.edu
Location: /pub/vmh/bin
FILE -rwxr-xr-x
23
495
Host pith.uoregon.edu
Location: /pub/Solaris2.x/bin
FILE -rwxr-xr-x
46952
Location: /pub/Sun4/bin
FILE -rwxr-xr-x
65536
Oct 27 12:09
newmail
Oct 27 12:10
newmail
Host ee.utah.edu
Location: /screen/bin
FILE -rwxr-xr-x
57344
Oct 11 1992
newmail
You can see that plaza.aarnet.edu.au (an educational facility in Australia—you
can tell because of the .au suffix), gum.isi.edu, venera.isi.edu, pith.uoregon.edu,
and ee.utah.edu all have one or more programs called newmail. There’s no way,
however, to ascertain from this listing whether it’s the program I’m seeking.
3. The same list can be produced in a more succinct format by using the -l command:
% archie -l newmail
19911220000000Z 15049
➥v25/newmail
19930709182600Z
104
19930709112600Z
104
19931027120900Z 46952
19931027121000Z 65536
19921011000000Z 57344
plaza.aarnet.edu.au /usenet/comp.sources.unix/
gum.isi.edu /share/pub/vmh/bin/newmail
venera.isi.edu /pub/vmh/bin/newmail
pith.uoregon.edu /pub/Solaris2.x/bin/newmail
pith.uoregon.edu /pub/Sun4/bin/newmail
ee.utah.edu /screen/bin/newmail
4. To search for all files that have something, anything, to do with mail (which is
going to generate a lot of output!), I can use the -s option:
% archie -s mail | more
Host plaza.aarnet.edu.au
Location: /usenet/comp.sources.unix/volume7
DIRECTORY drwxr-xr-x
512 Jan 16 1993
smail
Host metro.ucc.su.oz.au
Location: /pub/netinfo/sendmail
FILE -rw-r--r-12410 Jul 9 1992
Location: /pub/netinfo/sendmail/sendmail.mu
FILE -rw-r--r-15745 Oct 10 1990
sendmail.cf
sendmail.cf
Host brolga.cc.uq.oz.au
Location: /comp.sources.unix/volume7
512 Dec
1 1987
smail
Host cs.ubc.ca
Location: /mirror3/386BSD/386bsd-0.1/filesystem/etc
FILE -rw-r--r-17933 Jul 8 1992 sendmail.cf
--More-- _
It turns out that there are 95 matches, mostly comprising either sendmail.cf,
smail, or Rnmail .
23
Hour 23
496
5. The archie system also has a relatively limited database of descriptions called the
software description database, which you can check by directly connecting to a
remote archie system with telnet:
% telnet archie.unl.edu
Trying...
Connected to crcnis2.unl.edu.
SunOS UNIX (crcnis2)
login: archie
Last login: Wed Dec 15 10:47:17 from INS.INFONET.NET
SunOS Release 4.1.2 (CRCNIS2) #1: Wed Dec 16 12:10:12 EST 1992
too many archie users... try again later
As you can see, sometimes there are already too many people using the system for
you to log in and access their server.
6. I try an alternate site, archie.internic.net, the Internet Network Information
Center, and connect:
% telnet archie.internic.net
Trying...
Connected to ds.internic.net.
InterNIC Directory and Database Services
Welcome to InterNIC Directory and Database Services provided by AT&T.
These services are partially supported through a cooperative agreement
with the National Science Foundation.
First time users may login as guest with no password to receive help.
Your comments and suggestions for improvement are welcome, and can be
mailed to [email protected].
AT&T MAKES NO WARRANTY OR GUARANTEE, OR PROMISE, EXPRESS OR IMPLIED,
CONCERNING THE CONTENT OR ACCURACY OF THE DIRECTORY ENTRIES AND
DATABASE FILES STORED AND MAINTAINED BY AT&T. AT&T EXPRESSLY
DISCLAIMS AND EXCLUDES ALL EXPRESS WARANTIES AND IMPLIED WARRANTIES
OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.
SunOS UNIX (ds)
login: _
Any time you’re logging into an archie system, using the archie login is a good bet:
23
497
login: archie
*********************************************************************
Welcome to the InterNIC Directory and Database Server.
*********************************************************************
# Message of the day from the localhost Prospero server:
Welcome to Archie server for the
InterNIC Directory and Database Services.
# Bunyip Information Systems, 1993
# Terminal type set to `vt100 24 80’.
# èrase’ character is `^?’.
# `search’ (type string) has the value `sub’.
archie>
Now try the whatis command to search the software description database for mail
and generate a staggering number of matches.
JUST A MINUTE
In fact, there were many more matches than shown here. About 80
matches were made to Request for Comment documents available through
the Network Information Center.
archie> whatis mail
NMail
Novice Mail
answer
vacation(1) replacement. Answer mail while you’re away
batchmail
Convert batched news articles to a format suitable for
exchanging via electronic mail
bencode-bdecode
Binary-to-ASCII encoding scheme for mail
brkdig
Break mailing list digest into USENET messages
bsmtp
Batch SMTP (Simple Mail Transfer Protocol)
cfc
“Compile” sendmail.cf files into EASE language
cheap-fax
El-cheapo E-mail to Fax for sendmail
ck
Check mailboxes for new mail
ckmail
Check a user’s mail and report the “from” lines
clr-queue
Clean out the sendmail mail queue and send the results to
the system administrator
clr.queue
sendmail clean-up script
cms-unix
Transfer files (and files of mail data) between UNIX and
CMS (or MVS) systems
cobwebs
Check for old or unusually large mailboxes
cryptmail
Send and receive encrypted mail
deliver
Mail delivery agent which uses shell scripts as its
configuration files
distantbiff
Monitor distant mailboxes
dmail
Mail reading and sending program whic supports folders
and various methods of grouping messages by subject, address etc
dnamail
Send DECNET mail to/from a Sun running Sunlink/DNI
23
498
Hour 23
ease
Ease, a language for writing sendmail.cf files
elm
Elm (user agent) mail system
faces
Visual mail/print monitor
fido-usenet-gw
Implement a gateway between UUCP/Usenet/Mail and Fidonet
from
Mail summary generator
gate
Simple mail->news->mail gateway suite
gatech
GaTech Sendmail files
ida-sendmail
Enable sendmail to have direct access to dbm(3) files
and Sun Yellow Pages, separate envelope/header rewriting rulesets, and
multi-token class matches
junkmail
Delete outdated mail automatically
labels
Program to make mailing labels
lmail
A local mail delivery program
m
The more/mail/make/man thing
mail-s
Mail transmission with subject and suppression
mail.fixes
Patches to BSD4.2 mail (SysV mailx?)
maildigest
Construct a ARPA-style digest from a file of mail messages
mailias
“decode” mail aliases from your .mailrc and tell you who
things are going to
mailsplit
Send files and/or directories via electronic mail using
“tar”, “compress”, etc
mailwatcher
A Simple Mailwatcher
malias
Expand .mailrc aliases
mep102b
Mail Extensions Package. Handles things like
automatically tossing mail from people you don’t want to hear from, logging
incoming mail, and so on
mh-rn-interface
Method of interfacing the Rand MH mail handler with
the “rn” USENET news reading program
ml
Sort mail by Subject into separate files
mn
Mail summary/tally utility
mp
Mail pretty printer (aka mail->postscript)
mp23
A PostScript pretty printer for mail etc
mq
Display mail queue and “from” output
mq-from
PD replacements for mailq(1) and from(1) commands
msg
Screen oriented mail User agent
mush
Mail user’s shell
mverify
Mail alias/user verification
na-digest
Archive of mailings to NA distribution list (argonne)
netdata
Transfer data (and mail) between SysV and CMS
newsmail
Mail news articles to users automagically
nmail
Do UUCP mail routing using the output of the
pathalias(1) program
pc-mail-nfs
pc-mail over nfs
pcmail
Turn a PC into a (non-routing) UUCP node (DOS, PC unix)
pmdc
A “personal mail daemon” which filters mail much like GNU
Emacs does but without the overhead of Emacs and LISP
procmail
Mail processing package
returnmail
PD vacation(1). Answer your mail while you’re away
rmail-uucp
Domain Capable rmail for UUCP sites
round-robin
Mail round-robiner
savemap.nawk
A safe comp.mail.maps saver
sendmail-qref
A sendmail quick reference card
sendmail.ms
Sendmail reference card (troff -ms)
showhook.mh
MH Mail patch to allow actions when mail is read
23
sm-smtp
smail
smsmtp
SMAIL program
smtp_send
soundmail
sunmailwatch
tar-untar-mail
uumail
uumailclean
ux-maze
vacation
running sendmail
vmail
watch
wrap
xbiff
xmail
xmh
xwatch
archie>
499
Sendmail replacement for smail sites
A smart mailer and UUCP path router
SMTP server/client implementation for System V and the
SMTP SEND command for Sendmail
Sound mail
A mail watcher for SUNwindows
Sending tar(1) files through mail
Routing program to use the pathalias(1) database
Clean-up backlogged UUCP mail
UX-Maze Mail Based File Server
PD vacation(1) replacement for Berkeley systems not
Screen-based mail handler
A SysV program to display mail, time/date, and users on/off
Line wrap per for BIT/EARNnet mailings
Noification of new mail under X11
Mail front end for X11
X11 front end to the mh(1) mail agent
Replacement for xbiff and the mailbox widget(X11)
Now that the search is done, it’s time to log out:
archie> quit
# Bye.
%
7. To find where one of these programs is located, I can again use the local archie
program:
% archie mverify
Host ftp.germany.eu.net
Location: /pub/mail
512
Jul
7 15:15
mverify
It looks like the only host that has the program is in Germany!
With the capability to search the software description database, archie is a powerful
package for finding programs and information on the Internet. Remember that it’s
still limited by the ways that people might phrase or describe things as well as file-naming
conventions on each server.
Task 23.4: A Few Interesting telnet Sites
Two tasks on which I spend too much time are purchasing compact discs and books.
With the Internet, I can do both without leaving the privacy of my own computer
desk! Although these are commercial services, I illustrate them here to demonstrate the
incredible breadth of services available only on the Internet.
23
Hour 23
500
1. The first place to search is the Compact Disk Connection, an electronic record
store available through a system called Holonet:
% telnet orac.holonet.net
Trying...
Connected to orac.holonet.net.
HoloNet(SM) -- A service of IAT
HoloNet Member Name (Non-members type “guest”): _
To log in to the CD Connection, I enter cdc:
HoloNet Member Name (Non-members type “guest”): cdc
Last login: Wed Dec 15 13:04:18 from intrepid.ece.uc.
-----------------------HoloNet Services Gateway
-----------------------The HoloNet Services Gateway provides access to electronic services
through HoloNet. Use of this service is subject to HoloNet Terms
and Conditions.
The Compact Disc Connection is an independent service not affiliated
with Information Access Technologies, Inc.
Control-C to abort connect
Waiting for the Compact Disc Connection.....................
Connected to CD Connection.
Welcome to the
**
Compact Disc Connection
**
Dealing Exclusively in the Online Sale of
* Compact Discs *
and Featuring:
+=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=+
| FREE access from the Internet & from 75 Cities, NOW! |
+=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=+
- More Than 75,000 CDs Online -
23
501
- Discount Prices We accept VISA and MasterCard
*** CDC News *******************************************************
Want to know the status of your latest order? See the (C)heck order
status feature just added to the CDC Database Menu...
COMPACT DISC EUROPE, a Florida-based import company, is now online!
Looking for imports that aren’t in our catalog? Dial into their
online database of more than 100,000 imports from Europe and Japan
at: 408 730-8138. Any speed up to 9600, 8N1. Voice: 305 481-8984.
And speaking of imports, all
PHANTOM IMPORTS
ALEX IMPORTS
in our catalog are on sale now!
See (S)ales at the Main Menu...
*******************************************************************
** CDC Main Menu **
(C)Ds
(I)nformation
(N)umbers
(O)verseas
(A)ll-Music
(G)olden Ears
(S)ales
(F)ree CDs
(T)op Selling
(D)irectory
(P)ausing
(B)rief
(M)essage
(R)etrieve
(Q)uit
Enter CD database.
Display CDC policies and general information.
Display free modem access telephone numbers.
Display details of shipping to overseas destinations.
Display details of the All-Music Guide.
Display details of the Golden Ears Society.
Display details of current sales.
Display details of the free Adventures-in-Music.
Display Top Selling/Grammy Award Winning CDs.
Display the directory of CD labels and manufacturers.
Toggle pausing/no pausing of scrolling displays.
Toggle brief/full menu displays.
Leave a message to the management.
Retrieve messages from the management to you.
Sign off & hang up.
=> Your command: C
I want to search the CD database, so I enter C:
** CDC Database Menu **
(S)earch
(R)eview
(O)rder
(C)heck Status
(P)assword
(Q)uit
=> Your command: S
Search database and select CDs.
Review CDs you’ve selected.
Order CDs you’ve selected.
Check the status of your recent orders.
Change the password to your CDC account.
Return to the Main Menu.
23
Hour 23
502
I use the S key to request a search:
** CDC Search Menu **
(A)rtist
Search by artist or composer’s last name, e.g., Mozart, Dylan.
(S)ong
Search by song or track title, e.g., Star Spangled Banner.
(T)itle
Search by CD title, e.g., Woodstock.
(P)erformer Search for performers of classical music, e.g. Berlin Phil.
(M)anuf’er Search by manufacturer’s label, e.g., CBS
(N)umber
Search by manufacturer’s catalog number, e.g., 422 493-2.
(C)ategory Search by category of music, e.g., classical, rock.
(L)imits
Set limits for release date, music type, or Golden Ears ratings.
(1)-line
Toggle 1 or 2-line CD displays.
(E)xample
Display an example CD and an explanation of its components.
(Q)uit
Return to the Database Menu.
=> Your command: A
Then I search by artist:
** CDC Search by Artist **
Enter the first few letters of the artist/composer’s name (last, first),
or enter =STRING to search all positions in the artist name for STRING,
or press ENTER to repeat the previous search:
or enter a Q to quit: coltrane,j
MCA42001
$10.58
+COLTRANE*JOHN
MCA 9/88 1:07
AFRICA/BRASS VOL.1 & 2
PAB20101
$18.99
COLTRANE*JOHN
PABLO 12/93
AFRO BLUE IMPRESSIONS
CAP99175
$12.02
+COLTRANE*JOHN
CAPITOL 8/92
ART OF JOHN COLTRANE
oRi415
$10.50
PAUL
COLTRANE*JOHN
&ORIGINAL JAZZ CLASSICS
MCA5885
$10.79
+COLTRANE*JOHN
&MCA 5/88 :32
ATL1541
$10.74
+COLTRANE*JOHN
ATLANTIC 9/90
PAB2405417+COLTRANE*JOHN
$10.59
&PABLO 9/92
23
10/1
9/1 ***
BAHIA
**
2/90 WILBUR HARDEN, RED GARLAND,
BALLADS
8.7/3 **
BEST OF
***
BEST OF
**
:41
503
=> Enter a CD selector, a Q, or a ? for help: mca42001
=> Selected:
*MCA42001
$10.58
+COLTRANE*JOHN
MCA 9/88 1:07
=> 1 item(s) selected.
AFRICA/BRASS VOL.1 & 2
10/1
$10.58
=> Enter a CD selector, a Q, or a ? for help:_
That’s the CD I want. I easily could choose to buy it here and enter my VISA or
MasterCard number when prompted, and the disc would be mailed to me within a
week or so. For some cryptic reason, I decide I don’t need this album, and I quit
the program.
2. Now that I’ve exercised such self-restraint in avoiding the purchase of the Coltrane
album, how about buying a book or two? To connect to Book Stacks Unlimited in
Cleveland, Ohio, I use telnet books.com:
% telnet books.com
Trying...
Connected to books.com.
Book Stacks Unlimited, Inc.
Cleveland, Ohio USA
The On-Line Bookstore
Modem
: (216)861-0469
Internet : telnet books.com
Enter your FULL Name (e.g SALLY M. SMITH) :
I have an account, so follow me as I step through the book database, find a book,
and ensure that it’s the correct choice.
23
Hour 23
504
Type P to Pause, S to Stop listing
BOOK STACKS UNLIMITED, INC.
>>>> NEWS <<<<
1 - Internet Connection Is Now Available.
2 - Biblio-Tech -- The December Book.
3 - Helpful Information for Internet Callers.
NOTE: TELNET must be in character mode to echo keystrokes.
4 - Biblio-Tech. The Online Book Discussion Group.
99 - Recent Enhancements (Updated 10/23/93).
Press Enter to proceed to the Main Menu.
Type File # to View
<L>ist Files Again
<ENTER> To Exit :
****************************************
*
Book Stacks Unlimited, Inc.
*
*
MAIN MENU
*
****************************************
ook Store
<M>essages
<N>ews/Notes
<S>uggestions/Comments
<F>iles/Magazines
tilities
<H>elp
<G>oodBye
Command: B
23
505
****************************************************
*
The Book Store
*
*
273,481 Titles
*
****************************************************
<A>uthor Search
<R>eview Your Selections
<T>itle Search
<O>rder (when done)
<K>eyWord Title Search
<C>heck Order Status
SBN Search
<S>ubject Search \ Just Published
---------------------------------------------------revious Menu
<H>elp
<G>oodbye
Command: T
SEARCH DATABASE BY TITLE
Enter the first word(s) of the TITLE.
Omit leading ‘A’, ‘AN’,’THE’. The first few letters are enough.
Only the first 20 characters will be used.
<ENTER> Previous Menu, <?> Help, <ENTER>: tale of two cities
# TITLE
AUTHOR
PUB/BINDING/BK MARKS/PRICE
-----------------------------------------------------------------------1 A Tale of Two Cities
Dickens, Charles
07/90 Paperback S/O $ 7.95
2 A Tale of Two Cities
Dickens, Charles/Woodcock, George 06/85 Paperback S/O $ 4.95
Dickens, Charles
05/90 Paperback
12 $ 4.99
4 Tale of Two Cities
Dickens, Charles
08/91 Paperback
6 $ 2.95
Dickens, Charles
12/92 Paperback
12 $ 4.99
Dickens, Charles
09/89 Paperback
6 $ 2.50
7 Tale of Two Cities (Longman Classics, Stage 2)
Dickens, Charles
05/91 Paperback S/O $ 7.25
23
Hour 23
506
8 A Tale of Two Cities (World Classics)
Dickens, Charles
9 A Tale of Two Cities (Courage Classics)
Dickens, Charles
11/88 Paperback
S/O $
4.95
03/92 Hardcover
S/O $
5.98
<F>orward, ackward, revious Menu, <1-9> View Book # : 7
YOU HAVE SELECTED THE FOLLOWING TITLE:
Author
: Dickens, Charles
Title
: Tale of Two Cities (Longman Classics, Stage 2)
ISBN
:
Volume
:
Subject :
Dewey # :
Publisher:
Date Pub :
Binding :
Edition :
Bookmarks:
Price
:
0582030471
General Fiction
Addison Wesley (Longman)
05/91
Paperback
S/O
$ 7.25
How many copies would you like?, <ENTER> To Exit :
Again, I decide not to buy, and back out using the P (previous menu) option until I
can use G to say goodbye.
These are but two of hundreds of commercial services available on the Internet. One
of the best places to learn about the entire range of services available is to read Scott
Yanoff ’s List of Internet Services, available from anonymous FTP on csd4.csd.uwm.edu. If
you’re using fget, enter fget csd4.csd.uwm.edu:/pub/inet.services.txt.
This wraps up the tour of Internet navigational packages and Usenet. In these last two hours,
you have learned about archie and gopher, and you have traveled with me to library
computers in Australia and Venezuela, and even watched over my shoulder as I almost
ordered a John Coltrane CD and a recent edition of A Tale of Two Cities. And then there’s
the wonderful world of the Usenet!
23
The Internet is an amazing resource, and it’s growing dramatically each day. By the time you
read this, the size of the WAIS database list, the number of archie files in that database, and
the number of Usenet newsgroups will have expanded even further. If there’s an Information
Highway in your future, the Internet is most certainly going to be a key part of it, and you
can’t go wrong by spending some time learning more about it!
507
Workshop
in the next hour.
Key Terms
anonymous FTP A system set up to respond to ftp queries that does not require you to
have an account on the system.
bookmark A saved gopher menu item through which you easily can build your own
custom gopher information screens.
Request for Comment An official UNIX design specification, also known as an RFC.
search string The pattern specified in a search.
Questions
1. Use telnet and rlogin to log in to one of the sites shown in Table 23.3. You don’t
have an account, so drop the connection once you see a login: prompt.
2. Use ftp to connect to ftp.eff.org and see what files the Electronic Frontier
Foundation has made available to anonymous FTP users. Copy one onto your
system, and read through it to see what you think about the organization itself.
3. Using archie, find one or two archive sites that have a copy of the tin newsreading
program.
This completes your tour of the Internet. In your final hour, you are introduced to the basics
of C, the primary programming language for UNIX.
23
Hour
24
Programming in C
for UNIX
This hour introduces you to the basics of the C programming language. C is the
most commonly used language for programming UNIX systems. Other common languages are C++ and Perl, but C is the oldest, and many concepts are
derived from there. This chapter introduces a lot of concepts. I assume that you
have some basic math and programming skills, but even if you don’t, I
encourage you to skim through the material to learn about some of the
foundations of the UNIX system and how programmers can extend it in many
different directions. With any luck, you’ll have your interest piqued and decide
to learn how to get the computer to jump through hoops for you by writing your
own programs.
Goals for This Hour
In this hour, you learn to write your first program and about the following:
■ Basic data types and operators
■ Expressions
509
24
Hour 24
510
■
■
■
■
■
■
Conditional statements
Looping statements
Functions
Arrays
Pointers
Structures
First, you learn a simple program and how to make it run. After that, you learn the different
data types and how to manipulate the data. Control flow follows, where you learn how to
make your program execute alternate statements. You’ll wind up with some of the more
advanced topics in C programming.
Task 24.1: Your First Program
Historically, the first program written is called the “Hello, World” program because
it simply outputs that sentence. However, with the recent discovery of life on
meteoroids from Mars, and the suspicion that the oceans of Europa may also support life, we
should instead be greeting the universe.
The first program is actually very simple:
#include <stdio.h>
main()
{
printf(“Hello, universe!\n”);
}
Six simple lines; the first is an include line. This is a pre-processor instruction that tells the
compiler that when you build this program, it should include the contents of the named file
in addition to the code in this file. When included in <>, the compiler looks in the standard
directory, /usr/include. If the filename is quoted (as in #include “test.h”) it looks in the
current directory for the specified file.
JUST A MINUTE
24
A compiler is a special program found on any development system. It
reads in your program (source code), checks to make certain that it is
correct, and creates an executable program.
511
The file being included, stdio.h, is a header file that defines the standard input and output
functions. The .h is a naming convention indicating that the file is a header and is meant to
be used with the #include statements. Other common names used in C are .c for the source
file and .o for an intermediate object file.
The main is the program header. Every program, no matter how big or how small, must have
a main included. This must be followed by a curly brace, {, and any number of statements,
followed by a closing curly brace. This is the actual program. In our example, the main has
two parentheses following. C treats main as a normal function (described later). You can pass
arguments to main from a command line.
The statement is printf(“Hello, universe!\n”);. This calls the printf function, which
takes a string, possibly with some arguments, and places the contents of the string on the output.
Within the string, there is a pair of characters that you may not understand, \n. The backslash
is a C convention indicating that a special character follows. Table 24.1 lists the special C
characters.
Table 24.1. Special C characters.
Character
Meaning
\a
Bell character, which causes your terminal to beep
Backspace
New page
New line
Return
Tab
Vertical tab
Backslash
Question mark
Single quote
Double quote
Octal number (o is a digit between 0 and 7)
Hexadecimal number (h is a digit between 0 and f, where a is 10, b
is 11, c is 12, d is 13, e is 14, and f is 15)
\b
\f
\n
\r
\t
\v
\\
\?
\’
\”
\ooo
\xhh
This is a very simple program. Once you’ve written it, you need to save it to a file. Let’s call
that file hello.c. Once the file is saved, you can use the cc command to build the program.
% cc hello.c
24
Hour 24
512
This creates a file called a.out, which is an executable file. You can run a.out directly, or you
can use the mv command to give it a new name. Alternatively, you can add the -o option to
the cc command:
% cc -o hello hello.c
When you run the program, the output is
% hello
Hello, universe!
Of course, there are many different ways to write this program, but this is the most basic and
direct way to output a line of information to the screen.
This is only the beginning of learning C. This program only outputs a single string.
Task 24.2: Basic Data Types and Operators
Data in C programs can be kept in many forms. The basic data types are character,
integer, and floating point. These can be modified with standard operators, such as
addition and multiplication. Shell programs allow only for strings and arrays; with C, you
have more options.
Each C type is built from three basic data types. You can have a single character,
which is type char. The character is the amount of space needed to store a single
character. In languages that use the Roman alphabet, such as English, letters, digits, and
punctuation symbols are encoded in ASCII, which require seven bits to translate. Because
bytes are eight bits, and are a fairly universal data size, a character is allowed eight bits, or
one byte.
Other languages use all eight bits. French and Spanish need accents on vowels, and modifiers
on consonants. These character sets therefore use the full eight bits. Russian has an entirely
different character set, as does Greek, but because there are similarly small numbers of
characters, these char variables also are just one byte.
It is when one looks at non-European languages that the single byte causes a problem.
Japanese, Chinese, and other languages use a vastly larger number of characters in their script.
To accommodate this, in areas where it is needed, the char type is two, or even three, bytes.
These are sometimes called extended characters.
For the remainder of this book, English and ASCII characters are assumed, so one byte per
character is assumed.
24
The second type of variable is the integer variable. This is basically a counter, which can be
positive or negative. Usually, the integer is four bytes long, which gives it a range of
–2147483648 to 2147483647. All operations on an integer are integer operations, as
described in Hour 16, “Basic Shell Programming.” The type of an integer is int.
513
The third type of variable is a floating-point variable. This is how you would include
fractional, or irrational numbers. If you needed to list a radio station frequency, such as for
88.5 KQED, it would need a floating-point number. Results of uneven division can use
floating point. The type of a real number is float.
One weakness of floating point is rounding. Floating-point numbers have a limited
precision, the end result being a slow, gradual rounding error. If you perform many
calculations, this rounding error can grow to be significant.
JUST A MINUTE
A good example of this is your hand-held calculator. Enter the number 2,
then take the square root. You should see something like 1.414. Now,
square that number. On your calculator, you may see 1.999998. This is
the result of rounding error.
A lot of math still can be performed by integers. Programs that handle money often use
integers for the total number of cents, because this eliminates rounding error. Answers then
are presented as cents divided by 100 (dollars) and the remainder (cents).
These basic variables can be further modified. Integers can be modified with the adjectives
long and short. A long integer may use eight bytes and is architecture dependent. On an
eight-byte system, the range is then –9223372036854775808 to 9223372036854775807.
If you were to spell out that longest number, it would be nine quintillion, two hundred and
twenty-three quadrillion, three hundred and seventy-two trillion, thirty-six billion, eight
hundred and fifty-four million, seven hundred and seventy-five thousand, eight hundred and
seven. A formidable sum in any language.
A short integer is usually half the size of a regular integer but is also machine dependent. If
your machine has two-byte short integers, the range is –32768 to 32767.
Many C compilers allow you to omit the int when declaring short and
long integers.
JUST A MINUTE
The next pair of modifiers is signed and unsigned. Signed is usually assumed; so unsigned is
the important modifier.
Normally, an integer is considered to have a sign, so of the 32 bits available, one bit, usually
the first bit of the 32, acts as a sign flag. If the first bit is set to 1, you take the complement
of the number and treat it as the negative number of the same absolute value.
24
Hour 24
514
For example, if your bits are
00000000000000001011010010101111
the number you have is 46255. But, if your bits were
10000000000000001011010010101111
The number would be –2147437392.
You could set the integer to unsigned, and then the value would be 2147529903. This extends
the top end of the range of integers, at the cost of losing the ability to go negative.
Characters also can be unsigned. Because you can use characters for arithmetic (characters
are just bit patterns, as are integers), you also can make them unsigned. This is useful for times
when you need only a small range of values; an unsigned character has a range of 0 to 255.
JUST A MINUTE
A good example of this is in IP addresses. These addresses are a sequence of four numbers between 0 and 255. Many standard libraries
store these as arrays of four characters.
As with long and short, the int is not necessary for defining an unsigned integer.
The last modifier is for floating-point numbers. The additional type double is for a doubleprecision floating-point number. Most UNIX functions now default to double precision if
floating point is used.
All variable declarations must go after the opening curly brace. You can define any number
of variables, separated by commas, that you wish to use in a program. So, you’ll see:
int counter;
unsigned long mytaxes;
short myworth,your-worth;
double ratio;
char flag;
This declares six variables of five different types.
There are many different operators in UNIX that can be used on any variable. Table 24.2 lists
the unary operators first.
Table 24.2. Unary operators.
Operator
Meaning
++var
Increment the variable var before using it.
Increment var after using it.
var++
24
Operator
Meaning
--var
Decrement var before using it.
Decrement var after using it.
Negate the value of var.
Use the positive value.
Use the inverse value (for example, 10011100 becomes 01100011).
var--var
+var
!var
515
The increment and decrement operators add or subtract one unit from the value of var. The
unit for integers and characters is 1, for floating point, it is 1.0. This increment or decrement
can have greater meaning for pointers, where it steps to the next member of an array.
Table 24.3 lists the binary operators.
Table 24.3. Binary operators.
Operator
Meaning
a=b
a
a*b
a/b
a%b
a+b
a-b
a<>b
ab
a>=b
a==b
a&b
a^b
a|b
a&&b
a||b
is given the value of b
Multiply a and b
Divide a by b
The remainder of the division of a and b
The sum of a and b
The difference of a and b
The value of a shifted b bits to the left
The value of a shifted b bits to the right
The result of the comparison of a less than b
The result of the comparison
The bitwise AND operation of a and b
The bitwise XOR operation
The bitwise OR operation
The logical AND of a and b
The logical OR of a and b
24
Hour 24
516
Some of these may not make sense at first glance. Bitwise shifts are sometimes useful in
different UNIX functions. The same result can be obtained with multiplying the variable by
two for a one-bit shift left, or dividing by two for a one-bit shift right. For example, the
number 1234 is represented in binary as 0…10011010010. If you do a two-bit left shift,
you get 0…1001101001000. This is 4936. A one-bit right shift yields 0…1001101001. This
is 617.
The bitwise operators are more interesting. The result is a bit-by-bit comparison of equivalent
bits.
1
0
AND
OR
XOR
10
10
00
10
11
10
10
01
10
So, if you had two characters, 10110110 and 01101010, the results would be
AND
OR
XOR
JUST A MINUTE
00100010
11111110
11011100
These are useful for managing flags in a function. You can use AND and OR
operations to combine flags for the desired effect. Flags are single bits
that, when set, indicate a specific action must be performed. You can use
separate variables for each flag, but it is cleaner to include them all in a
single variable.
You assign a value to a variable with a simple equals sign.
{
int a;
float b;
char c;
a=1;
b=3.1415;
c=’c’;
Note that a single character must be enclosed in single quotes. Later in this chapter, when you
look at strings, those need to be enclosed in double quotes. Each of the preceding operators
creates an expression. Assignment statements are also expressions. An assignment statement
must have a single variable on the left ‘“”and any expression or value on the right. This can
create some unusual assignments, illustrated here:
24
A=b=c=d=1;
A=((b+c)*d)<=((f/e)<<2);
517
Because an assignment is an expression, you can assign multiple variables the same value, as
the first line illustrates. There, the variables A, b, c, and d are all assigned the value 1. The
second line is more complicated. First, you add b and c, and then multiply by d. Then, you
divide f by e, and do a left bitwise shift of two places. Then, you compare the two values, and
if the first is less than or equal to the second, you assign 1 to A; otherwise, you assign 0.
There is a shorthand for many assignments. If you have a=a+4; as a statement, this can be
reduced to a+=4;. Any operation where the results of the operation are assigned to one of the
operands can be so abbreviated. So, you can perform a bitwise OR with a|=b;, and you can
multiply with a*=b;.
Now you know the basic statements, where you can assign the results of an expression
to a variable. You have also learned the different variable types, and a bit about their use.
Task 24.3: Conditional Statements
Conditional statements enable you to take different actions as the result of the
evaluation of an expression.
Any manipulation of a variable in C is considered an expression; even the simple
assignment is an expression that can be evaluated to a specific value. Furthermore,
the results of any expression can be used in a subsequent expression.
You can use an if statement to test an expression and perform an action. The syntax is
if (expr) statement-block;
You optionally can have an else clause:
else statement-block;
A statement-block is either a single statement followed by a semicolon or a listing of statements
surrounded by curly braces.
So, if you wanted to test a specific value to see if it is greater than 1000000, you easily could
do this:
if (value>1000000) printf(“I am rich\n”);
The else statement can give an alternate answer:
if (value>1000000) printf(“I am rich\n”);
else printf(“I have to go to work today.\n”);
Each expression tested evaluates to a true or false value. C uses 0 for false and 1 for true.
However, C also has expanded this to allow any non-zero value to be true. Because zero also
is the value NULL, as used when reads fail, you can have
if (fgets(stdin,buffer,1024))
{
/* Manipulate the input */
}
24
Hour 24
518
is a means of reading a line of input, stdin is the standard input file (usually your
terminal), and buffer is an array of characters.
fgets
One weakness of the if statement is that, with else, the parsing can be tricky. Consider
this case:
if (a<b) if (c<d) { /* Do something */ }
else { /* Do something else */ }
This is a perfectly valid statement in C but is very ambiguous. Is the else clause to be executed
if a=d? Or is it executed only if a>=b? This depends on the compiler, but it usually
will default to the highest valid unmatched condition, that is, if a>=b. If you meant the former,
though, you can force that result by putting the second if, with the else, in a statementblock:
if (a<b)
{
if (c<d)
{
/* Do something */
}
else
{
/* Do something else */
}
}
You similarly can force the other interpretation with:
if (a<b)
{
if (c<d)
{
/* Do something */
}
}
else
{
}
An alternative forcing is to have an empty block with {} symbols and nothing within:
if (a<b)
if (c<d)
{
/* Do something */
}
else {}
else
{
}
24
519
There is a very convenient shorthand for simple if statements in C. This is the ?: notation.
It takes three expressions and executes the second if the first is true; otherwise, it executes the
third. Think of this as condition?true-action:false-action.
A useful form of this is the “plural” statement:
printf(“%d apple%c”,count,(count>1)?”s”:””);
Here, the comparison is used in the print statement. If there is more than one apple
(count>1), the string “s” is appended to apple. You’ll find this shorthand in many C
programs.
A second type of conditional is the
multiple options. The syntax is
switch
statement. This allows for the evaluation of
switch(expression)
{
case const: statements;
case const: statements;
...
default:
}
The expression can evaluate to any value, and it is then compared with the constant values
of each case.
If this expression evaluates to a string, using strings in cases won’t work; a
returned string is just a memory address.
JUST A MINUTE
If you were testing input for one of three values as an answer to a question that you output
to confirm an action from the user, you’d use this sequence of instructions:
switch(c=getchar())
{
case ‘y’:
case ‘Y’:printf(“The answer is Yes.\n”); break;
case ‘n’:
case ‘N’:printf(“Alas, the answer is no.\n”); break;
case ‘m’:
case ‘M’:printf(“The answer is maybe?\n”);break;
default:printf(“I do not understand the answer.\n”);
}
Note the use of break; after each statement. This indicates that you have finished the
execution of this switch block and want to drop to the subsequent code section. Normally,
the case statements are just labels, so the program resumes execution from that location.
24
Hour 24
520
This drop-through can have its uses; if you want to do the same thing for every valid case but
also want to say something in a special case, you can do it:
switch(c=getchar())
{
case ‘Y’:printf(“No need to shout.\n”);
case ‘y’:printf(“That’s an affirmative.\n”);break;
}
Both y and Y will get the message That’s
an affirmative. ,
but only Y will see No
need to
shout.
Conditional statements add the power to execute alternative statements.
Task 24.4: Looping Statements
There are times when you want to be repetitive. The for, while, and do statements
are ideal in this case.
The first looping statement is the while loop. These loops test a condition, and while
the condition is true, they execute the following code. The syntax is
while (expr) statement
The statement can be a null statement, if desired. To step through white space in an array
of characters, you might use
while (str[i++]==’ ‘);
While is particularly useful for an indeterminate loop. You just keep executing the statement
until the condition is false.
A special case is the infinite loop. You will see this in certain types of programs that wait for
events:
while(1)
{
/* Get event */
/* Action */
}
The only way to end this program is when the action calls for an exit because the condition
1 is always true
The second kind of loop is the for loop. It takes three expressions, an initialization, a test,
and an increment. As long as the test is true, the statement block is executed. The syntax is
for(expr1;expr2;expr3) statement;
24
521
This is exactly the same as:
expr1;
while(expr2)
{
statement;
expr3;
}
The choice of loop is up to you.
For loops are particularly useful for stepping through arrays or in any situation where the test
is related to an initialization.
For example, you could count the number of characters in an array with:
for (i=0; c[i] != 0; i++);
You need to remember that an array ends with a character with the value 0, so any real
character is true.
None of the conditions needs to be present. A loop for(;;) is an infinite loop.
The third loop is the do loop. It is the same as the while loop, except the test comes after
executing the statement block:
do
statement
while (expr);
This forces at least one execution of the statement.
You can exit a loop regardless of the condition with a break; statement. This causes you to
execute the first statement following the loop. Breaks can occur anywhere in the loop.
You can restart the loop with continue;. In a while loop, continue forces the testing of the
condition, then another run through the loop. With a for loop, continue forces the
increment expression to be run, then the test, before starting the statement block. With do,
the test is executed, then the loop may be restarted.
There are three basic types of loops in C, each usable in different circumstances.
Task 24.5: Functions
You can create a more modular program with functions, instead of attempting to
include all your statements under main. If you have a piece of code that needs to be
executed in different areas of the program, by making it a function, you have a smaller
program with the same power.
24
Hour 24
522
There are two types of functions available in C. Strictly speaking, only one is a
function; the other is called a macro replacement. Functions start off with a header,
this defines the function and its arguments. The complete function syntax is
type name(arglist) { statements }
The type can be any; it defines the type of variable returned by the function. In addition to
the three basic types, there is a fourth, special type, called void. This is used if there is to be
no return.
The argument list is a comma-separated list of variables, with a type specified. Even if
multiple variables are of the same type, they each need the specifier.
The statements can be any statements at all. In this case, I have a converter from centigrade
to Fahrenheit.
double fahrenheit(double ctemp)
{
return (ctemp*1.8+32.0);
}
This is actually a fairly simple function. The variable ctemp is multiplied by 1.8 (9/5), and
has 32 added. You then can include the function in a program.
main()
{
printf(“The temperature in Fahrenheit when it is”);
printf(“ 23.3 centigrade is %4.1f\n”,
fahrenheit(23.3));
}
Turns out to be a rather balmy 73.9 degrees. Another, more interesting function is one that
returns the difference between a centigrade and a Fahrenheit temperature in either centigrade
or Fahrenheit:
double temperature-difference(double ctemp,double ftemp,char corf)
{
double cftemp;
double difference;
cftemp=ctemp*1.8+32;
difference=(cftemp>ftemp)?cftemp-ftemp:ftemp-cftemp;
return ((corf)?(difference-32)/1.8:difference);
}
This is a fairly complicated function. It takes three arguments, a centigrade temperature, a
Fahrenheit temperature, and a flag. If the flag is true, the function returns the difference in
centigrade. If false, the return is in Fahrenheit.
The function first converts the centigrade temperature into Fahrenheit. It then tests the
difference; the test makes sure the difference is always positive. Then, it returns the difference,
converted back to centigrade, if needed.
24
523
The macro is not really a function, but it can be used like one. A macro is a command to the
C compiler to replace one piece of text with another. It must occur before the replacement
in the file, and it often looks like:
#define
#define
CENTIGRADE
FAHRENHEIT
1
0
So, any place where CENTIGRADE is used, 1 is replaced. For the preceding difference function,
that is quite convenient and much easier to read, too:
diff=temperature-difference(centigrade,fahrenheit,CENTIGRADE);
This is more clear than:
diff=temperature-difference(centigrade,fahrenheit,1);
You can specify an argument to the macro, too. Any number can be specified. In this example,
I’ve replaced the conversion of centigrade to Fahrenheit with a macro:
#define
fahrenheit(X)
X*1.8+32
Now, when a program that calls Fahrenheit is compiled, instead of including a function call,
this text—with the X replaced with a value—is substituted and compiled.
You have just had a brief introduction to functions. These are C constructs to group
statements that are repeated and which can take arguments and return data.
Task 24.6: Arrays
You can associate a group of data items by using arrays.
Arrays are the means in C where you can declare a list of related objects. The most
common type of array is comprised of characters, but arrays of integers and floats are
not uncommon.
To declare an array, after you declare the name, follow it by the number of elements:
char string[100];
This declares that string is an array of 100 characters. The name is not chosen randomly,
though. A string, in C, is just an array of characters and can be manipulated as such.
When you access arrays, the first element of the array is always 0. For users of other languages,
you may have seen the first element start with 1. That is not the case here. So, an array of size
100 has elements numbered 0 to 99.
24
Hour 24
524
Another good example of an array is this prime-number builder:
#include <stdio.h>
main()
{
int primes[25];
int counter=1;
int start=3;
int prime;
int i;
primes[0]=2;
printf(“2”);
while (counter<25)
{
prime=1;
for(i=0;i<counter&′i++)
if (!(start%primes[i])) prime=0;
if (prime)
{
printf(“ %d”,start);
primes[counter++]=start;
}
start++;
}
printf(“\n”);
}
This program builds an array of prime numbers and then uses it to determine further primes.
The output is
% primes
2 3 5 7 11 13 17 19 23 29 31 37 41 43 47 53 59 61 67 71 73 79 83 89 97x
In this task, you were introduced to arrays. Arrays provide a means of relating
common data. The most common use is to create a string, which is just an array of
characters.
Task 24.7: Pointers
Each variable in a C program has a given location in memory. That location, or
address, can be assigned to a pointer.
To understand pointers, you first need to understand how memory is organized.
Every time you declare a variable, a piece of memory is allocated in a given size, and
is labeled with your variable name. Any time you access that name, the value of the data at
that location is provided.
So, when you declare
int i;
24
525
a piece of memory the size of an integer is allocated for you to use. In the program, this piece
of memory is tracked with the name ‘i’.
Suppose, though, you want to know the address of that piece of memory. You can take it with
the unary & operator. The value assigned must be a pointer:
int i;
int *pointertoint;
The asterisk indicates that the variable pointertoint is not an integer, but an address of an
integer. Later, you can assign it the address of i with:
pointertoint= &i;
You then can access that value with (*pointertoint).
This is particularly useful in functions. When you call a function, the arguments you pass to
the function remain intact, even after you execute the function. This is called “call by value.”
So, if you had a function swap and passed it a and b, you’d see this:
void swap(int a,int b)
{
int c;
c=a;
a=b;
b=c;
return;
}
main()
{
int a,b;
a=4;
b=2;
printf(“%d %d\n”,a,b);
swap(a,b);
}
Here is an example of the output:
%
%
4
4
cc -o swapper swapper.c
swapper
2
2
The swap function did not swap the values. If you want them to change, you need to pass an
address and do some pointer arithmetic:
void swap(int *a, int *b)
{
int c;
24
Hour 24
526
c=(*a);
(*a)=(*b);
(*b)=c;
return;
}
main()
{
int a,b;
a=4;
b=2;
swap(&a,&b);
}
Here is an example of the output of the modified program:
%
%
4
2
%
cc -o swapper swapper.c
swapper
2
4
By passing the addresses, you can assign the new value to the address in memory of the
variables. This is called “call by reference.”
Interestingly, pointers and arrays are very closely related. When you declare an array, the
memory is allocated for you for all the elements, contiguously. Then, you access this array
by an index. You can assign the address of the first member to a pointer as usual:
int array[10];
int *point;
point=&array[0];
Now, when you increment the pointer, it increments by the size of the array members. So,
point+1 now points to the second array member. In fact, *(point+i) is the same as array[i] .
The declaration of an array just declares a pointer to that location, and array offsets are
calculated as increments to that pointer.
Pointers are an interesting way to access and pass data in a manageable fashion
between functions. Instead of using the value in memory, a pointer is the address of
the memory. This allows the passing of addresses between functions and the modification
of values within functions.
Task 24.8: Structures
Structures are groupings of unlike data types into a single object. This object can be
referenced directly, and the members of the structure can be similarly referenced.
24
527
Suppose you wanted to manage a student’s academic record. You’d expect to find
in there a name, an identification number, and perhaps a GPA. You easily could
create this structure:
struct academic {
char name[100];
int id;
double gpa;
};
The structure created here is called academic, and it has three fields: a name, which is a string;
id, an integer; and gpa, a double-precision floating-point number. Now, I can use this in a
program:
Char *getname(void)
{
static char buffer[1024];
printf(“Enter a name: “);
fgets(buffer,1024,stdin);
while(strlen(buffer)<1)
{
printf(“No name entered, please try again: “);
fgets(buffer,1024,stdin);
}
return buffer;
}
double getgpa(void)
{
static float gpa;
printf(“Enter the GPA: “);
scanf(“%f”,&gpa);
while ((gpa<0.0)||(gpa>4.0))
{
printf(“The GPA must be between 0 and 4: “);
scanf(“%f”,&gpa);
}
return gpa;
}
main()
{
struct academic {
char name[100];
int id;
double gpa;
} students[20];
int i;
for(i=0;i<20;i++)
{
students[i].id=i;
strcpy(students[i].name,getname());
students[i].gpa=getgpa();
}
}
24
Hour 24
528
This program creates 20 student records. You can assume that getname and getgpa prompt
for name and GPA information, so they are entered manually. You can access members of
a structure with the . notation, and you can make arrays of structures.
Structures also can have pointers, but the means of accessing the members of a structure are
slightly different. Although you could, perhaps, use (*students).id, the mechanism
students->id looks a bit better.
The -> symbol tells the program that you are using a pointer to reference a part of memory,
and you need the specific offset into memory to find a field.
Another type of a structure is the union. In this case, only one member of a union can be
accessed at any time. Structures can be viewed as a collection of fields, all included in the data.
Unions are means of providing access to only one piece of data, but that data can be
interpreted in different fashions.
Unions are not commonly used in early programs, but you may see a union when examining
programs.
You now have a grasp of how data in a C program can be related by structures and
how to declare a structure.
Summary
This completes a basic walk-through of the C programming language. C is an enormous
topic, and there are many books on the subject. The definitive book, if you want to learn more
about C, is The C Programming Language by the original authors of C, Brian Kernighan and
Dennis Ritchie. Of all the programming books James has acquired and read over the years,
his C programming language book is the only book always available to him; his copy is a dogeared third printing of the first edition published back in 1978.
Where To Go Next
This marks the end of your journey. In 24 hours, you’ve learned considerably more about
UNIX than most people ever learn, and I hope you’ve had fun along the way. Just like any
large body of information, particularly one that evolves daily, there’s still a lot more to learn.
To get from here to there, I have a few suggestions.
To learn more about the Internet, I again recommend the enjoyable and valuable Navigating
the Internet by Mark Gibbs and Richard Smith. I’ve read it a couple of times, and each time
I find something new and amusing. If nothing else, you will learn what the words aalii and
zymurgy mean, which is a potential boon next time you play Scrabble.
24
529
To learn more about C programming, read Teach Yourself C in 21 Days by Peter Aitken and
Bradley Jones and the official language definition, The C Programming Language by Brian
Kernighan and Dennis Ritchie. Your UNIX vendor also should have supplied information
on C programming tools available with your system.
If you want to become a true UNIX power user, I recommend UNIX Unleashed, from Sams
Publishing. It is stuffed full of interesting and valuable information about the many UNIX
commands on your system.
There are some valuable documents available on the Internet, too: Scott Yanoff has an
Internet Services List that can be quite informative. Visit it online for yourself at http://
www.spectracom.com/islist/. In addition to the list of Usenet newsgroups that you can
access with the findgroup alias shown earlier, there are thousands of electronic mailing lists.
You can obtain a very large listing of all groups by obtaining the file rtfm.mit.edu:/pub/
usenet/news.answers/mail/mailing-lists. Finally, a list of some of the more fun information servers on the Internet can be obtained as cerebus.cor.epa.gov:/pub/bigfun.
Finally, don’t forget that your UNIX system has lots of documentation and information, and
most of it’s online! For any command you find yourself using frequently, the man page entry
might well show you new ways to combine things, to work with starting options and files,
and more. Always look for an EXAMPLES section at the end of the document, and don’t forget
that you can print it by using man cmd | lpr at the command line.
Have fun, and enjoy UNIX! It’s the most powerful operating system you can work with, and
it’s only as easy or complex as you let it be. Tame the beast and study what’s in this book and
other books on the subject, and you’ll grow to appreciate the system.
Visit the official Web site for this book, too, to get any last-minute updates and pointers to
tons more useful and interesting UNIX information online. It’s at http://www.intuitive.com/
tyu24.
Workshop
topics presented in this chapter.
Key Terms
bitwise operator An operator that works directly on the bits, without changing neighboring bits.
bitwise shift Changing the location of the bits in memory.
compiler A program that takes source code and makes it executable.
24
Hour 24
530
expression A C language construct that has a value.
extended characters A means of displaying non-Latin characters, such as Japanese,
Chinese, or Arabic characters.
Questions
1.
2.
3.
4.
24
How would you create the galaxy? How would you greet the solar system?
Which types are best for these variables: Social Security number? Eye color? Name?
Which loop is better for stepping through elements in an array?
How would you build a structure for a driver’s license record?
Glossary
Glossary
531
Glossary
532
absolute filename Any filename that begins with a leading slash (/); these always uniquely
describe a single file in the file system.
access permission The set of accesses (read, write, and execute) allowed for each of the
account name This is the official one-word name by which the UNIX system knows you:
mine is taylor. (See also account in Hour 1.)
account This is the official one-word name by which the UNIX system knows you. Mine
is taylor.
addressing commands The set of vi commands that enable you to specify what type of
object you want to work with. The d commands serve as an example: dw means delete word,
and db means delete the previous word.
anonymous FTP A system set up to respond to ftp queries that does not require you to
have an account on the system.
arguments Not any type of domestic dispute, arguments are the set of options and
filenames specified to UNIX commands. When you use a command such as vi test.c, all
words other than the command name itself (vi) are arguments, or parameters to the program.
basename The closest directory name. For example, the basename of /usr/home/taylor is
taylor.
binary A file format that is intended for the computer to work with directly rather than for
humans to peruse. See also executable.
bitwise operator An operator that works directly on the bits, without changing neighboring bits.
bitwise shift Changing the location of the bits in memory.
blind carbon copy
recipient.
An exact copy of a message, sent without the awareness of the original
block special device A device driver that controls block-oriented peripherals. A hard disk,
for example, is a peripheral that works by reading and writing blocks of information (as
distinguished from a character special device). See also character special device.
block At its most fundamental, a block is like a sheet of information in the virtual notebook
that represents the disk: a disk is typically composed of many tens, or hundreds, of thousands
of blocks of information, each 512 bytes in size. See also i-node to learn more about how disks
are structured in UNIX.
bookmark A saved gopher menu item through which you easily can build your own
custom gopher information screens.
Glossary
bookmarks A listing of favorite sites for quick retrieval.
browser A program designed to load hypertext pages and follow hyperlinks.
buffer
An area of the screen used to edit a file in emacs.
carbon copy An exact copy of a message sent to other people. Each recipient can see the
names of all other recipients on the distribution list.
character special device A device driver that controls a character-oriented peripheral.
Your keyboard and display are both character-oriented devices, sending and displaying
information on a character-by-character basis. See also block special device.
colon commands The
manipulation.
vi
commands that begin with a colon, usually used for file
column-first order When you have a list of items that are listed in columns and span
multiple lines, column-first order is a sorting strategy in which items are sorted so that the
items are in alphabetical order down the first column and continuing at the top of the second
column, then the third column, and so on. The alternative strategy is row-first order.
command alias A shorthand command mapping, with which you can define new
command names that are aliases of other commands or sequences of commands. This is
helpful for renaming commands so that you can remember them, or for having certain flags
added by default.
command block A list of one or more shell commands that are grouped in a conditional
or looping statement.
command history A mechanism the shell uses to remember what commands you have
entered already, and to enable you to repeat them without having to type the entire command
again.
command mode The mode in which you can manage your document; this includes the
capability to change text, rearrange it, and delete it.
command number The unique number by which the shell indexes all commands. You can
place this number in your prompt using \! and use it with the history mechanism as
!command-number .
command Each program in UNIX is also known as a command: the two words are
interchangeable.
compiler A compiler is a program that takes source code and makes it executable.
conditional expression This is an expression that returns either true or false.
533
534
Glossary
control-key notation A notational convention in UNIX that denotes the use of a control
key. There are three common conventions: Ctrl-C, ^c, and C-C all denote the Control-c
character, produced by pressing the Control key (labeled Control or Ctrl on your keyboard)
and, while holding it down, pressing the c key.
control number A unique number that the C shell assigns to each background job for easy
reference and for using with other commands, such as fg and kill.
core dump The image of a command when it executed improperly.
current job The job that is currently running on the terminal and keyboard (it’s the
program you’re actually running and working within).
determinant loop A loop where the number of times the loop is run can be known before
starting the loop.
device driver All peripherals attached to the computer are called devices in UNIX, and
each has a control program always associated with it, called a device driver. Examples are the
device drivers for the display, keyboard, mouse, and all hard disks.
directory A type of UNIX file used to group other files. Files and directories can be placed
inside other directories, to build a hierarchical system.
directory separator character On a hierarchical file system, there must be some way to
specify which items are directories and which is the actual filename itself. This becomes
particularly true when you’re working with absolute filenames. In UNIX, the directory
separator character is the slash (/), so a filename like /tmp/testme is easily interpreted as a file
called testme in a directory called tmp.
domain name UNIX systems on the Internet, or any other network, are assigned a domain
within which they exist. This is typically the company (for example, sun.com for Sun
Microsystems) or institution (for example, lsu.edu for Louisiana State University). The
domain name is always the entire host address, except the host name itself. (See also host
name.)
dot A shorthand notation for the current directory.
dot dot A shorthand notation for the directory one level higher up in the hierarchical file
system from the current location.
dot file A configuration file used by one or more programs. These files are called dot files
because the first letter of the filename is a dot, as in .profile or .login. Because they’re dot
files, the ls command doesn’t list them by default, making them also hidden files in UNIX.
See also hidden file.
dynamic linking Although most UNIX systems require all necessary utilities and library
routines (such as the routines for reading information from the keyboard and displaying it
Glossary
to the screen) to be plugged into a program when it’s built (known in UNIX parlance as static
linking), some of the more sophisticated systems can delay this inclusion until you actually
need to run the program. In this case, the utilities and libraries are linked when you start the
program, and this is called dynamic linking.
e-mail
Electronically transmitted and received mail or messages.
errant process A process that is not performing the job you expected it to perform.
escape sequence An unprintable sequence of characters that usually specifies that your
terminal take a specific action, such as clearing the screen.
exclusion set A set of characters that the pattern must not contain.
executable A file that has been set up so that UNIX can run it as a program. This is also
shorthand for a binary file. You also sometimes see the phrase binary executable, which is the
same thing! See also binary.
expression A C language construct that had a value. A command that returns a value.
extended characters A means of displaying non-Latin characters, such as Japanese,
Chinese, or Arabic characters.
file-creation mask When files are created in UNIX, they inherit a default set of access
permissions. These defaults are under the control of the user and are known as the filecreation mask.
file redirection Most UNIX programs expect to read their input from the user (that is,
standard input) and write their output to the screen (standard output). By use of file
redirection, however, input can come from a previously created file, and output can be saved
to a file instead of being displayed on the screen.
filter Filters are a particular type of UNIX program that expects to work either with file
redirection or as part of a pipeline. These programs read input from standard input, write
output to standard output, and often don’t have any starting arguments.
flags Arguments given to a UNIX command that are intended to alter its behavior are
called flags. They’re always prefaced by a single dash. As an example, the command line
ls -l /tmp has ls as the command itself, -l as the flag to the command, and /tmp as the
argument.
flow control The protocol used by your computer and terminal to make sure that neither
outpaces the other during data transmission.
foreground job A synonym for current job.
heuristic A set of well-defined steps or a procedure for accomplishing a specific task.
hidden file By default, the UNIX file-listing command ls shows only files whose first letter
535
Glossary
536
isn’t a dot (that is, those files that aren’t dot files). All dot files, therefore, are hidden files, and
you can safely ignore them without any problems. Later, you learn how to view these hidden
files. See also dot file.
home directory This is your private directory, and is also where you start out when you
log in to the system.
host name UNIX computers all have unique names assigned by the local administration
team. The computers I use are limbo, well, netcom, and mentor, for example. Enter hostname
to see what your system is called.
hyperlinks Specifications within a document that include instructions for loading a
different document.
i-list See i-node.
i-node The UNIX file system is like a huge notebook full of sheets of information. Each
file is like an index tab, indicating where the file starts in the notebook and how many sheets
are used. The tabs are called i-nodes, and the list of tabs (the index to the notebook) is the
i-list.
inclusion range
A range of characters that a pattern must include.
indeterminant loop A loop where the number of times the loop is run is not known before
starting the loop.
insert mode The vi mode that lets you enter text directly into a file. The i command starts
the insert mode, and Escape exits it.
interactive program An interactive UNIX application is one that expects the user to enter
information and then responds as appropriate. The ls command is not interactive, but the
more program, which displays text a screenful at a time, is interactive.
job A synonym for process.
job control A mechanism for managing the various programs that are running. Job control
enables you to push programs into the background and pull them back into the foreground
as desired.
kernel The underlying core of the UNIX operating system itself. This is akin to the
concrete foundation under a modern skyscraper.
key bindings The emacs term for key mapping.
key mapping A facility that enables you to map any key to a specific action.
kill
Terminate a process.
left rooted Patterns that must occur at the beginning of a line.
Glossary
login shell The shell you use, by default, when you log in to the system.
login A synonym for account name, this also can refer to the actual process of connecting
to the UNIX system and entering your account name and password to your account.
loop This is a sequence of commands that are repeatedly executed while a condition is true.
mail folder A file containing one or more e-mail messages.
mail header The To:, From:, Subject:, and other lines at the very beginning of an e-mail
message. All lines up to the first blank line are considered headers.
mailbox A synonym for mail folder.
major number For device drivers, the major number identifies the specific type of device
in use to the operating system. This is more easily remembered as the device ID number.
man page Each standard UNIX command comes with some basic online documentation
that describes its function. This online documentation for a command is called a man page.
Usually, the man page lists the command-line flags and some error conditions.
Meta key Analogous to a Control key, this is labeled either Meta or Alt on your keyboard.
minor number Once the device driver is identified to the operating system by its major
number, the address of the device in the computer itself (that is, which card slot a peripheral
card is plugged into) is indicated by its minor number.
modal A modal program has multiple environments, or modes, that offer different
capabilities. In a modal program, the Return key, for example, might do different things,
depending on which mode you were in.
mode A shorthand way of saying permissions mode.
modeless A modeless program always interprets a key the same way, regardless of what the
user is doing.
multitasking A multitasking computer is one that actually can run more than one
program, or task, at a time. By contrast, most personal computers lock you into a single
program that you must exit before you launch another.
multiuser Computers intended to have more than a single person working on them
simultaneously are designed to support multiple users, hence the term multiuser. By contrast,
personal computers are almost always single-user because someone else can’t be running a
program or editing a file while you are using the computer for your own work.
named emacs command A command in emacs that requires you to type its name, like
query-replace , rather than a command key or two.
null character Each character in UNIX has a specific value, and any character with a
537
Glossary
538
numeric value of zero is known as a null or null character.
password entry For each account on the UNIX system, there is an entry in the account
database known as the password file. This also contains an encrypted copy of the account
password. This set of information for an individual account is known as the password entry.
pathname UNIX is split into a wide variety of different directories and subdirectories,
often across multiple hard disks and even multiple computers. So that the system needn’t
search laboriously through the entire mess each time you request a program, the set of
directories you reference are stored as your search path, and the location of any specific
command is known as its pathname.
permission strings The string that represents the access permissions.
permissions mode The set of accesses (read, write, and execute) allowed for each of the
This is a synonym for access permission.
pipeline
A series of UNIX commands chained by |, the pipe character.
preference file These are what dot files (hidden files) really are: they contain your
individual preferences for many of the UNIX commands you use.
preserve Ensure that a message doesn’t move out of your incoming mailbox even though
you’ve read it.
print job name The unique name assigned to a print job by the lpr or lp command.
print queue The queue, or list, in which all print jobs are placed for processing by the
specific printer.
process A program stopped or running within the UNIX operating system. Also known
as a job.
recursive command A command that repeatedly invokes itself.
regular expressions A convenient notation for specifying complex patterns. Notable
special characters are ^ to match the beginning of the line and $ to match the end of the line.
relative filename Any filename that does not begin with a slash (/) is a filename whose exact
meaning depends on where you are in the file system. For example, the file test might exist
in both your home directory and in the root directory: /test is an absolute filename and leaves
no question which version is being used, but test could refer to either copy, depending on
your current directory.
replace mode
the file.
A mode of
vi
in which any characters you type replace those already in
Glossary
Request for Comment An official UNIX design specification, also known as an RFC.
root directory
The directory at the very top of the file system hierarchy, also known as slash.
row-first order In contrast to column-first order, this is when items are sorted in rows so
that the first item of each column in a row is in alphabetical order from left to right, then the
second line contains the next set of items, and so on.
search path A list of directories used to find a command. When a user enters a command
ls, the shell looks in each directory in the search path to find a file ls, either until it is found
or the list is exhausted.
search string The pattern specified in a search.
shell To interact with UNIX, you type in commands to the command-line interpreter,
which is known in UNIX as the shell, or command shell. It’s the underlying environment in
which you work with the UNIX system.
shell alias Most UNIX shells have a convenient way for you to create abbreviations for
commonly used commands or series of commands, known as shell aliases. For example, if I
always found myself typing ls -CF, an alias can let me type just ls and have the shell
automatically add the -CF flags each time.
signals Special messages that can be sent to stopped or running processes.
slash The root directory.
standard error This is the same as standard output, but you can re-direct standard error
to a different location than standard output.
standard input UNIX programs always default to reading information from the user by
reading the keyboard and watching what’s typed. With file redirection, input can come from
a file, and with pipelines, input can be the result of a previous UNIX command.
standard output When processing information, UNIX programs default to displaying the
output on the screen itself, also known as standard output. With file redirection, output can
easily be saved to a file; with pipelines, output can be sent to other programs.
starting flag Parameters that you specify on the command line when you invoke the
program.
stop a job Stop the running program without terminating it.
subshell A shell other than the login shell.
surfing A style of interacting with the World Wide Web, usually for pleasure, where you
follow hyperlinks from Web site to Web site.
539
Glossary
540
symbolic link A file that contains a pointer to another file rather than contents of its own.
This can also be a directory that points to another directory rather than having files of its own.
A useful way to have multiple names for a single program or allow multiple people to share
a single copy of a file.
tilde command A command beginning with ~ in Berkeley Mail or the Elm Mail System.
transpose case
undelete
Switch uppercase letters to lowercase or lowercase to uppercase.
Restore a deleted message to its original state.
URL The specification for a document on the World Wide Web. Usually, it includes a
protocol, machine name, and filename.
user environment A set of values that describe the user’s current location and modify the
behavior of commands.
user ID A synonym for account name.
variables These are names to label data that may change during the execution of a program.
wedged process A process that is stuck in memory and can’t free up its resources even
though it has ceased running. This is rare, but annoying.
wildcards Special characters that are interpreted by the UNIX shell or other programs to
have meaning other than the letter itself. For example, * is a shell wildcard and creates a
pattern that matches zero or more characters. Prefaced with a particular letter, X—X* —this
shell pattern will match all files beginning with X.
working directory The directory where the user is working.
World Wide Web A collection of sites that provide hypertext documents on the Internet.
XON/XOFF A particular type of flow control. The receiving end can send an XON (delay
transmission) character until it’s ready for more information, when it sends an XOFF
(resume transmission).
zero-length variable
zombie
A variable that does not have a value assigned to it.
A terminated process that has not been cleaned up by the parent process.
INDEX
Symbols
“ (double quote), 124
‘ (single quote), 124
nl command, 157
! (exclamation point), domain
names, 28
! command
emacs editor, 296
vi editor, 270-277
awk command (creating shell
scripts), 274-277
fmt command (tightening
paragraph lines), 273-274
ls-CF command output,
adding to files, 271-272
paragraphs, assigning to
UNIX commands, 272-273
!! command (C shell history
commands), 333
!$ command (C shell history
commands), 333
!* command (C shell history
commands), 333
!* notation (aliases), 340
!n command (C shell history
commands), 333
$ command (vi editor), 219
$ motion command (vi editor),
250-251
$ notation (egrep command), 175
$ prompt, 24
% (percent sign), 24
& symbol (moving background
processes), 367
( command (vi editor), 268
) command (vi editor), 268
* (asterisk) wildcard, 129
filename wildcards, 162-164
echo command, 162
limiting number of matches,
162-164
regular expressions, 167
+ (vi editor), 234
+= notation, 193
-? flag, 14, 16
-1 flag (ls command), 68, 77, 152
. notation (egrep command), 175
./ls command, 56
/ (filenames), 46
/ command (vi editor), 268
/ search command (vi editor),
231-232
= command, 141
= command (rn program), 474
? command (vi editor), 232-233
? wildcard (filename wildcards),
162-164
echo command, 162
@ (filenames), 46
[^xy] notation (egrep command),
175
\” character (C programming
language), 511
\’ character (C programming
language), 511
542
\? character
\? character (C programming
language), 511
\\ character (C programming
language), 511
^ notation (egrep command), 175
{ command (vi editor), 270
~ command (vi editor), 228
24-hour time, 31
A
\a character (C programming
language), 511
a command (vi editor), 216-219
A command (emacs help system),
298
-a flag, 54
du command, 80-81
ls command, 68, 77
dot files, listing, 67
ps command, 368
(a) notation (egrep command), 175
â^b command (C shell history
commands), 333
a|b notation (egrep command), 175
abbreviations (.exrc file)
defining, 262
expanding, 262-263
listing abbreviations in effect, 263
absolute filenames, 51-52, 60
absolute numeric values
(permissions), 98
access permission, 84
account names, 22, 40
accounts, 6
passwords, changing, 25
acronym database, 451-452
addresses, decoding, 445
addressing commands (vi editor),
243
adjust command, 273
-ag flag (ps command), 372-373
agenda program (hidden files), 55
alias command, 334-335
man pages, scrolling, 14
alias feature, 124
aliases, 67, 335-337
!* notation, 340
cd (customizing prompts),
339-340
cd command, 337
chdir command, 337
connecting to accounts on other
systems, 336-337
defining, 335
printer-specific print
commands, 386
setprompt (customizing
prompts), 339-340
setprompt command, 337
alt newsgroups, 467
AltaVista Web site, 436-437
American Standard Code for
Information Interchange
(ASCII), 129
American Telephone and
Telegraph (AT&T), 3
anonymous FTP, 507
anonymous ftp capabilities (Web
sites), 483
AnswerBook, 17
append command (vi editor),
216-218
Apple Computer FTP archive,
489-490
apropos command, 8
man page, 12
archie program, 493-499
archie servers, 494
interacting with archie databases,
493-494
Internet Network Information
Center, 496
logging in, 496-497
whatis command (mail
searches), 497-499
program searches, 494-495
-l command, 495
-s command, 495
software description
database, 496
archive random library file, 132
archives, finding with archie
system, 493-499
archie servers, 494
Center, 496-499
searching specific programs,
494-495
software description database,
496
arguments, 35
command-line arguments
assigning variables, 350
for command, 358
arithmetic functions (variables),
350-351
ARPAnet, 480
arrays (C programming language),
523-524
arrow keys, mapping (vi editor),
265
ASCII, 129
ascii command (ftp program), 483
assigning variables
command-line arguments, 350
read command, 349
assignment statements (C
programming language), 516-517
asterisk wildcard, 129
echo command, 162
at sign (@), filenames, 46
-atime n flag (find command), 398
autoexec.bat (DOS file), 50
awk command, 177, 188-195, 370
awk program files, creating, 194
BEGIN pattern, 192
END pattern, 192-194
error messages, 195
-f flag, 188
-Fc flag, 188
if-then condition, 194
interacting with ! command
(vi editor), 274-277
login shells, 189
modifying output of ls command,
192
NF variable, 190-191
NR variable, 191
print command, 188-189
sorting data, 190
tables, building, 194-195
awk filters, 366
awk script (finding local printers),
381-382
-ax flag (ps command), 373
B
\b character (C programming
language), 511
^b command (vi editor), 219
b command, 141
vi editor, 219
C shell
B command
emacs help system, 298
vi editor, 219
-b flag
sort command, 151
test command, 354
b key command (vi editor), 211
B programming language, 5
background processes, 365-368
& symbol (moving automatically), 367
files, processing, 366
input/output, 368
job ID, 365
logging out, 377
starting automatically, 366
z command, 367
Backspace command (vi editor),
219
base-8 numbering system, 38
basename, 339, 344
bc command (infix calculator),
36-38
options, 37
BEGIN pattern (awk command),
192
Berkeley Fast File System, 5
Berkeley Mail, 411-416
command options, 412
delete command, 415-416
forwarding messages, 414
headers command, 414-415
index numbers, 413-414
message display, 412-413
message headers, 414
quit command, 416
responding to messages, 414
save command, 415
groups of messages, 416
sending mail, 417-420
~h command, 419
~p command, 418
~r command, 418
tilde commands, 417-418
Berkeley Systems Design Web
site, 429
Berkeley UNIX systems (mail
folders), 129
bg command (background
processes), 365, 367
bin directory, 46
binary, 60
permission conventions, 99
binary command (ftp program),
483
binary comparisons (file system),
354
binary format, 46
binary operators (C programming
language), 515
bitwise operators, 529
C programming language, 516
bitwise shift, 529
blind carbon copy, 424
block special devices, 133, 143
blocks, 66, 85
Book Stacks Unlimited telnet site,
503-506
Bookmark menu (Netscape
Navigator), 431
bookmarks, 507
gopher program, 454, 456-457
boot file, 49
Bourne shell, 2, 306
$ prompt, 24
alias feature, 124
Bourne Again shell, 307
jsh shell, 307
restricted sh shell, 307
break; statement (exiting loops),
521
browsers, see Netscape Navigator
BSD, 2
buffers, 303
emacs editor, 300-302
moving between buffers, 301
reading files into buffers, 300
bye command (ftp program), 483
C
C command
change commands, 246-248
c command (change commands),
246
ranges of text, changing, 249-252
-C flag (ls command), 68, 77
-c flag
grep command, 172, 174
more program, 139
test command, 354
uniq command, 150
\c notation (egrep command), 175
c notation (egrep command), 175
543
c* notation (egrep command), 175
c+ notation (egrep command), 175
c? notation (egrep command), 175
C programming language
arrays, 523-524
compiler, 510
conditional statements, 517-520
if statement, 517-519
switch statement, 519-520
functions, 521-523
macro replacements, 523
looping statements, 520-521
do loop, 521
for loop, 520-521
while loop, 520
pointers, 524-526
printf function, 511
sample program, 510-512
special C characters, 511
structures, 526-528
union structures, 528
variables, 512-517
binary operators, 515
character sets, 512
character variable, 512, 514
floating-point variables, 513
integer variable, 512-514
unary operators, 514-515
values, assigning, 516-517
C shell, 2, 306
aliases, 335-337
!* notation, 340
cd command, 337
chdir command, 337
connecting to accounts on
other systems, 336-337
command-alias mechanism,
333-335
defining an alias, 335
DOS commands,
re-creating, 334
general format, 333
ls command flags, 333-334
configuration files, 317-320
.cshrc file, 319-320
.login file, 317-319
environment, 313-317
variables, 314-317
history commands, 333
544
C shell
history list, 327-333
building, 327-328
command numbers, 328
csh history mechanism,
330-331
echo command, 330
file wildcards, 329-330
filename searches, 331-332
repeating commands,
328-329
history mechanism, 324-327
reviewing histories, 324
set history command, 326
shell parameters, checking
status, 325
starting, 325
w command output, 325
jobs (processes)
fg command, 362, 364
running in background,
365-368
screen-oriented programs,
363-364
stopping, 362-365
search path, 91
setting custom prompts, 338-340
cd alias, 339-340
cwd variable, 338
setprompt alias, 339-340
values, 338
tcsh shell, 307
C shell environment modification
command, 320
C shell unalias command, 178
C-a command (emacs editor), 289
C-b command (emacs editor),
287, 289
C-c command (emacs help system),
298
C-d command (emacs editor), 293
C-d command (emacs help system),
298
C-e command (emacs editor), 289
C-f command (emacs editor),
287, 289
C-h command (emacs editor), 287
C-k command (emacs editor), 293
C-n command (emacs editor),
286, 289
C-n command (emacs help system),
298
C-p command (emacs editor), 289
C-v command (emacs editor), 289
C-w command (emacs help
system), 298
C-x [ command (emacs editor), 289
C-x ] command (emacs editor), 289
C-x Delete command (emacs
editor), 293
C-x u command (emacs editor),
293
cal command, 33
calculators, 36
bc infix, 36, 38
dc postfix, 38-39
RPN, 36
sine, 38
calendar, displaying, 33
cancel command (print queues),
391
cap mail, 411
carbon copy, 424
case, transposing (vi editor), 228
case command, 357
cat program
numbering file lines (-n flag),
153-154
pipelines, 138
-s flag, 136-137
-v flag, 136
viewing file contents, 136-139
catching signals, 376
cd alias (customizing prompts),
339-340
cd command, 58
aliases, 337
cd dir command (ftp program),
483
cdrom directory, 49
-CF flag (ls command), 486
change command, 246-252
lines, changing contents, 247-248
ranges of text, changing, 249-252
change shell (chsh) command,
310, 312-313
character range (filename
wildcards), 164-166
character sets, 512
character special devices, 133, 143
character variable, 512
unsigned variable, 514
chdir command (aliases), 337
chgrp command, 109-110
chmod command
modifying permissions, 96-98
numeric permissions strings, 102
setting new permissions, 98-102
symbolic notation, 96
chmod function, 90
chown command, 108-109
chsh (change shell) command,
124, 312-313
Clear key, mapping, 264
clock
24-hour time, 31
displaying time, 33
close command (ftp program), 483
colon commands (vi editor),
236-241, 243
:e command, 241
:n command, 240
:r command, 238-239
:w command, 238
column-first order, 85
comma-separated lists, 68
command alias command, 320
command aliases, 306-307, 321
command history, 321
command shells, 306-307
command mode (vi editor), 243
command number, 344
command prompts, see prompts
command shells, see shells
command-alias mechanisms
(shells), 333-335
DOS commands, re-creating, 334
general format, 333
command-blocks, 355, 360
command-line arguments
assigning variables, 350
for command, 358
command-line interfaces, 2
command-line systems, 6
command.com (DOS file), 50
commands
! (emacs editor), 296
! (vi editor), 270-277
awk command (creating shell
scripts), 274-277
ls-CF command output,
adding to files, 271-272
paragraphs, assigning to
!! (C shell history commands),
333
commands
!$ (C shell history commands),
333
!* (C shell history commands),
333
!n (C shell history commands),
333
$ (vi editor), 219
lines of text, changing,
250-251
( (vi editor), 268
) (vi editor), 268
./ls, 56
/ (vi editor), 268
/ search command (vi editor),
231-232
=, 141
rn program, 474
? (vi editor), 232-233
{ (vi editor), 270
a (vi editor), 219
A (emacs help system), 298
â^b (C shell history commands),
333
addressing commands (vi editor),
243
adjust, 273
alias, 334-335
append (vi editor), 216-218
apropos, 8
arguments, 35
ascii (ftp program), 483
awk, 177, 188-195, 370
awk program files, creating,
194
BEGIN pattern, 192
error messages, 195
-f flag, 188
-Fc flag, 188
interacting with ! command
login shells, 189
modifying output of ls
command, 192
NR variable, 191
sorting data, 190
^b (vi editor), 219
b, 141
vi editor, 219
B
vi editor, 219
Backspace (vi editor), 219
bc (infix calculator), 36, 38
bg (background processes),
365, 367
binary (ftp program), 483
bye (ftp program), 483
C (change commands), 246
lines, changing contents,
247-248
c (change commands), 246
ranges of text, changing,
249-252
C (emacs help system), 298
C shell environment modification, 320
C shell unalias, 178
C-a (emacs editor), 289
C-b (emacs editor), 287, 289
C-c (emacs help system), 298
C-d (emacs help system), 298
C-e (emacs editor), 289
C-f (emacs editor), 287, 289
C-h command (emacs editor),
287
C-k (emacs editor), 293
C-n (emacs editor), 286, 289
C-n (emacs help system), 298
C-p (emacs editor), 289
C-v (emacs editor), 289
C-w (emacs help system), 298
C-x [ (emacs editor), 289
C-x ] (emacs editor), 289
C-x Delete (emacs editor), 293
C-x u (emacs editor), 293
cal, 33
cancel (print queues), 391
case, 357
cd, 58
aliases, 337
cd dir (ftp program), 483
change, 246-252
247-248
249-252
change shell (chsh), 312-313
chdir (aliases), 337
chgrp, 109-110
chmod
modifying permissions, 96-98
symbolic notation, 96
545
chown, 108-109
chsh, 124
close (ftp program), 483
colon commands (vi editor),
236-241, 243
:e command, 241
:n command, 240
:r command, 238-239
:w command, 238
command alias, 320
Control-l, 141
cp, 7, 116-117
^d (vi editor), 219
d, 141
vi editor, 220-221
D (vi editor), 220
date, 33
dc (postfix calculator), 38-39
dc -l math, 47
dd (vi editor), 224-225
delete (mailx), 415-416
df, 82-83
dir (ftp program), 484
remote directory files, listing,
484-485
disk (man page), 10
du
-a flag, 80-81
disk space usage, checking,
79-81
multi-letter flags, 81
dw (vi editor), 222-223
:e (vi editor), 241
:e filename (vi editor), 236
echo, 58, 155
filename wildcard characters,
162
history lists, 330
egrep, 175-176
notations, 175
emacs motion commands, 289
entering in more program, 142
env, 57, 313-314
Escape (vi editor), 219
exit, 24
expr (arithmetic functions),
350-351
:f, 141
^f (vi editor), 219
F (emacs help system), 298
fdformat (man page), 12
fg
starting jobs, 365
stopping jobs, 362, 364
546
commands
fgrep, 176-179
awk command, 177
excluding words from lists,
178-179
wrongwords file, 176-178
file
core dumps, 131
determining accuracy of, 128
file permissions, 130
identifying file types,
128-130
mail folders, 129
symbolic links, 131
file-redirection, 146-147
file-related commands (emacs
editor), 299-302
find, 398-403
core files, removing, 402-403
-exec flag, 402-403
flags, 398
-l flag, 404
listing files/directories,
398-399
-mtime n flag, 400
-name flag, 399-400
-type specifier, 401-402
xargs command, 403-405
fmt (tightening paragraph lines),
273-274
for, 358
^g (vi editor), 256
g (rn program), 472
G (vi editor), 229-231
get
ftp program, 484
transferring login file, 486
grep, 172-174
-c flag, 172, 174
character searches, 169-170
general form, 173
-i flag, 172-173
inclusion range searches, 168
-l flag, 172, 174
letter searches, 169
-n flag, 172, 174
punctuation sequence
searches, 170-171
specific word searches, 168
~h (mailx), 419
h, 141
vi editor, 219
headers (mailx), 414-415
help, 14, 16-17
history, 329
C shell, 333
exclamation points, 327
repeating commands, 327
i (vi editor), 219
I (emacs help system), 298
id, 30
if, 355-356
insert (vi editor), 216-217
j (vi editor), 219
jobs, 369
k (vi editor), 219
K (emacs help system), 298
kill, 370, 374-377
hang-up signal (SIGHUP),
376
kill signal (SIGKILL), 376
repeating, 376
signals, 374-375
-l (archie program), 495
l (vi editor), 219
L (emacs help system), 298
lcd dir (ftp program), 484
lp
flags, 384
printing files, 384-387
lpinfo, 386-387
lpq (print queues), 391
lpr
flags, 384
lprm (print queues), 391
ls, see ls command
ls -C -F, 60
ls -C -F /, 46
M (emacs help system), 298
M-< (emacs editor), 289
M-> (emacs editor), 288, 289
M-a (emacs editor), 289
M-b (emacs editor), 289
M-d (emacs editor), 293
M-Delete (emacs editor), 293
M-e (emacs editor), 286, 289
M-f (emacs editor), 289
M-k (emacs editor), 293
M-v (emacs editor), 289
map (vi editor)
arrow keys, mapping, 265
control characters, 264
saving key mappings, 265
mesg, enabling messages,
408-409
mget
ftp program, 484
wildcard patterns, specifying,
486
mkdir, 114-116
man page, 9
-More- prompt, 141
mput (ftp program), 484
mv, 116
moving files, 118-119
renaming files, 119-120
:n (vi editor), 236, 240
n, 141
emacs editor, 296
vi editor, 232-234
n[Return], 141
named emacs command,
294, 303
netinfo, 445
nf, 141
nl
-bp pattern-matching
option, 156
command flag format, 155
default settings, 156
numbering file lines, 153-157
quotation marks, 157
-s flag, 156-157
ns, 141
numeric repeat prefixes
O (vi editor), 215-216, 219
o (vi editor), 214-215, 219
open (ftp program), 484
overview, 2
~p (mailx), 418
passwd, 25
/pattern, 141
pr
-f flag, 390
flags, 388
formatting print jobs,
387-391
piping output to lpr
command, 390-391
print, 188-189
printenv, 57
prompt (ftp program), 484
config.sys
ps, 368-374
-ag flag, 372-373
-ax flag, 373
flags, 368-369
Sequent workstation
output, 374
Sun workstation output, 374
-u flag, 372
punctuation in, 28
put (ftp program), 484
pwd, 58
ftp program, 484
:q (vi editor), 220, 237
:q! (vi editor), 230, 237
q, 141
emacs editor, 296
quit (mailx), 416
:r (vi editor), 238-239
~r (mailx), 418
R (replace commands), 246
r (replace commands), 246
characters, changing,
248-249
:r filename (vi editor), 237
read (assigning variables), 349
rehash, 91
printers script, 382
replace, 246-252
248-249
Return (vi editor), 219
rlogin
logging in to remote systems,
480-481
logging out of remote
systems, 481
rm, 121, 123
precautions with, 123-125
rmdir, 120-121
-s (archie program), 495
S (emacs help system), 298
save (mailx), 415
groups of commands, 416
search-and-replace (vi editor),
257-260
changing all occurrences,
259-260
words, changing, 258-259
sed, 179-184
assigning paragraphs to
deleting lines in streams,
181-182
g flag, 180
multiple sed commands,
separating, 181
prefixes, adding/removing,
184
regular expressions, 182-183
rewriting output of who
command, 181
substituting patterns, 180
semicolon, 123
set (configuration options), 319
set history, 326
:set number (vi editor), 256
setprompt, aliases, 337
several on one command
line, 123
sort
file information, sorting,
150-153
filenames, sorting alphabetically, 152
flags, 151
lines of a file, sorting, 152
man page, 13
[Space], 141
spacing in, 28
spell, 178-179
stty (configuration options), 318
stty tostop, 365
substitute, 179-180
T (emacs help system), 298
tee (re-routing pipelines),
196-197
telnet
connecting to remote archie
system, 496
remote system connections,
481-483
test, 351-355
file system status tests,
353-355
numeric comparisons, 352
string comparisons, 353
test operators, 352
unary flags (file systems),
353-354
time, 33
touch, 91
files, creating, 78-79
tty, 408-409
û (vi editor), 220
u (vi editor), 221-222, 252
umask, 104, 106-107
547
uncompress, 84
undo (emacs editor), 292-293
uniq, 149-150
users, 31
v, 141
WAIS, databases, 451
V (emacs help system), 298
vi (starting vi editor), 201
:w (vi editor), 220, 237-238
W
vi editor, 220
w, 32
vi editor, 220
:w filename (vi editor), 237
wc -l /etc/magic, 128
whatis (archie program), 497-499
who, 31
rewriting output with sed
command, 181
whoami, 28
wq (vi editor), 229
write, 409-411
x (vi editor), 220
xargs, 403-405
y (emacs editor), 296
z (background processes), 367
see also programs
communicating with users
enabling messages (mesg
command), 408-409
write command, 409-411
see also e-mail
Compact Disk Connection telnet
site, 500-503
comparison functions (test
command), 351-355
file system status tests, 353-355
test operators, 352
compilers, 510, 529
complex expressions, 175-176
compress program, 83-84
computing sums, 193
concentric circles of access, 98
conditional expressions, 360
case command, 357
if command, 355-356
conditional statements, 517-520
config.sys (DOS file), 50
548
configuration files
configuration files
C shell, 317-320
printers, /etc/printcap file,
380-381
connections
device drivers, tty command,
408-409
remote Internet sites, 480-483
logging in, 481
logging out (rlogin
command), 481
rlogin command, 480
telnet command, 481-483
continue; statement (restarting
loops), 521
control key notation, 136, 143
control number, 377
Control-f command (vi editor),
209-210
control-key commands (vi editor),
208-209
b key (moving around files word
by word), 211
Control-f command, 209-210
Control-u command, 210
w key (moving around files word
by word), 211
Control-l command, 141
Control-u command
(vi editor), 210
copying
files (cp command), 116-117
Internet files (ftp program),
483-493
anonymous ftp capabilities,
483
Apple Computer FTP
archive, 489-490
binary files, 490-491
to computer screen, 489
FTP archives, 491-492
ftp commands, 483-484
MIT AI Laboratory
(anonymous FTP archive),
487-488
484-486
starting ftp, 484
core dumps, 131, 143
core files, removing with find
command, 402-403
counting words/lines (wc
program), 147-149
cp command, 7, 116-117
csh (command shell), 306
configuration files, 317-320
csh history mechanism, 330-331
.cshrc file
apropos command, 8
C shell configuration files,
319-320
displaying contents, 137
viewing with more program, 139
-ctime n flag (find command), 398
Ctrl-c, 37
Ctrl-d, 37
CTSS operating system, 4
current job, 377
cursor control keys (vi editor),
205-208
cursors, command prompts, 24
customizing
Netscape Navigator (General
Preferences), 437-440
Appearance tab, 437
Applications tab, 438
Fonts tab, 437
Helpers tab, 439
Images tab, 439
prompts, 338-340
cd alias, 339-340
cwd variable, 338
values, 338
D
^d command (vi editor), 219
d command, 141
vi editor, 220-221
D command (vi editor), 220
-d flag, 93
lp command, 385
ls command, 78
listing directories, 71
more program, 139
sort command, 151
test command, 354
uniq command, 150
d suffix, 447
databases
acronyms, 451-452
recipes, 452-453
searching with WAIS, 449-454
software description database
(archie program), 496
v command, 451
date, displaying, 33
date command, 33
dc -l math command, 47
dc command (postfix calculator),
38-39
dd command (vi editor), 224-225
DEC
PDP-11, 4
PDP-7, 4
default environment variables,
314-316
HOME, 314
LOGNAME, 316
MAIL, 316
NAME, 316
PATH, 315
SHELL, 314
TERM, 314-315
USER, 315
default login shells, viewing,
308-309
defining
abbreviations (.exrc file), 262
aliases, 355
delete command (mailx), 415-416
deleting
directories (rmdir command),
120-121
files (rm command), 121, 123
precautions, 123-125
text
dd command, 224-225
dw command, 222-223
editing typos, 225-227
u command, 221-222
vi editor, 220-229
deletion commands (emacs
editor), 293
determinate loops, 357, 360
/dev directory, 47, 132-133
device drivers, 47, 60
block special devices, 133, 143
character special devices, 133,
143
Elm Mail System
major number, 144
minor number, 144
tty command, 408-409
df command, 82-83
diag directory, 49
dir command (ftp program), 484
484-485
directories, 60
absolute filenames, 51-52
bin, 46
cdrom, 49
changing (cd command), 58
copying (cp command), 117
diag, 49
etc, 47
file system
/dev directory, 132-133
/lib directory, 132
core dumps, 131
symbolic links, 131
top level, 130-131
group
changing (chgrp command),
109-110
identifying, 107-108
home, 49
/lib directory, 47, 132
listing (ls command), 69-71
lost+found, 48
mnt, 48
net, 49
new (mkdir command), 114-116
owner
changing (chown command),
108-109
pcfs, 49
relative filenames, 51-52
removing (rmdir command),
120-121
renaming (mv command), 120
root, 45
searching, 56
sizes, indicating (ls command),
65-66
blocks occupied, 66
sys, 48
tftpboot, 49
tmp, 48
usr, 48
directory permissions, 93-96
execute, 93-94
execute-only, 94
modifying (chmod command),
96-98
read, 93-94
write-only, 95
directory separator characters,
50, 61
directory trees, listing recursively
(ls command), 73-74
disk command (man page), 10
disk space
available disk space, checking
(df command), 82-83
usage, checking (du command),
79-81
do loop, 521
domain names, 29, 40
! (exclamation point), 28
domain naming, 477
domain-based naming scheme, 29
domains (e-mail addresses),
444-445
top-level domains, 444
DOS commands, re-creating, 334
dot files, 61
listing, 67
see also hidden files
double quote (“), 124
-dptr flag (lp command), 384
drctry permissions, setting new
(chmod command), 98, 100, 102
du command
-a flag, 80-81
disk space usage, checking, 79-81
multi-letter flags, 81
Duke Basketball Report Web
site, 435
dw command (vi editor), 222-223
dynamic linking, 48, 61
DYNIX, 9, 22
dynix file, 49
E
:e command (vi editor), 241
:e filename command (vi editor),
236
-e flag (test command), 354
549
e-mail, 424, 426
addresses, 444-445
decoding, 445
d suffix, 447
Elm Mail System, 420-423
reading messages, 421
replying to messages,
422-423
starting, 420-421
finger command, 446
From: line (basic notations), 445
gopher files, electronic mailing
of, 458
mailx
~h command, 419
message headers, 414
~p command, 418
quit command, 416
~r command (reading in
files), 418
reading messages, 411-416
save command, 415-416
netinfo command, 445
sending to Internet users,
444-446
addresses, 444-445
echo command, 58, 155
filename wildcard characters, 162
history lists, 330
echoing passwords, 23
editing typos (vi editor), 225-227
tilde (~) command, 228
EDITOR environment
variable, 332
-ef flag (test command), 354
egrep command, 175-176
notations, 175
electronic books, accessing (gopher
program), 458-459
electronic mail, see e-mail
reading messages, 421
replying to messages, 422-423
starting, 420-421
550
elm program
elm program (hidden files), 54
emacs editor, 201
buffers, 300-302
reading files into, 300
deleting text, 289-293
restoring deleted text, 293
sentences, removing, 291-292
undo command, 292-293
words, removing, 290-291
deletion commands, 293
file-related commands, 299-302
GNU EMACS, 283
Help options (C-h command),
287
help system, 297-299
options, 297-298
motion commands, 289
moving around in files, 285-289
C-b command, 287
C-f command, 287
C-n command, 286
M-> command, 288
M-e command, 286
notation for indicating
commands, 282-283
quiting, 283
saving file changes, 288
searching and replacing, 294-297
backward searches, 294-295
query-replace feature,
294-296
starting, 282-284
undo command, 292-293
END pattern (awk command),
192-194
env command, 57, 313-314
environment variables, 314-317
EDITOR, 332
EXINIT, 317
HOME, 314
LOGNAME, 316
MAIL, 316
NAME, 316
PATH, 315
PRINTER, 383, 385
SHELL, 314
TERM, 314-315
USER, 315
environment variables (rn
program)
RNINIT, 469
SUBJLINE, 474
-eq flag (test command), 352
errant processes, 377
error messages (awk command),
195
Escape command (vi editor), 219
escape sequence, 279
etc directory, 47
/etc/passwd file, 142-143
/etc/printcap file, 380-381
finding specified printers, 383
/etc/shells file, 312
exclamation point (domain
names), 28
exclusion set, 185
-exec flags (find command),
398, 402-403
executables, 46, 61
execute permissions
directory, 93-94
file, 90-91
execute-only permission (directory), 94
EXINIT variable (environment
variables), 317
exit command, 24
expanding abbreviations (.exrc
file), 262-263
expr command (arithmetic
functions), 350-351
expressions, 360, 530
C programming language,
517-520
conditional, see conditional
expressions
.exrc file (vi editor), 260-264
defining abbreviations, 262
expanding abbreviations,
262-263
listing abbreviations in effect, 263
extended characters, 530
external media (mnt and sys
directories), 48
F
\f character (C programming
language), 511
:f command, 141
^f command (vi editor), 219
F command (emacs help system),
298
-F flag (ls command), 68, 78
filename suffixes, appending, 67
-f flag
awk command, 188
pr command, 388, 390
sort command, 151
test command, 354
-Fc flag (awk command), 188
fdformat command (man page), 12
fg command
starting jobs, 365
stopping jobs, 362, 364
fgrep command, 176-179
awk command, 177
178-179
file command
core dumps, 131
determining accuracy of, 128
identifying file types, 128-130
mail folders, 129
symbolic links, 131
file permissions, 88, 90-93
binary equivalents, 100
file command, 130
96-98
numeric equivalents, 101-103
setting new (chmod command),
98-102
file redirection, 146-147, 158
viewing files, 146-147
file system, 45-46, 130-133
bin directory, 46
binary comparisons, 354
cd command, 58
cdrom directory, 49
core dumps, 131
files
diag directory, 49
directory separator characters, 50
echo command, 58
env command, 57
etc directory, 47
hidden files, 52-53
home directory, 49
HOME variable, 57
lost+found directory, 48
Macintosh, differences, 50
mnt directory, 48
net directory, 49
PATH variable, 57
PC, differences, 50
pcfs directory, 49
pwd command, 58
status tests (test command),
353-355
unary flags, 353-354
symbolic links, 131
sys directory, 48
tftpboot directory, 49
tmp directory, 48
top level, 130-131
usr directory, 48
file systems
Berkeley Fast File System, 5
Thompson, 4
file types (identifying with file
command), 128-130
mail folders, 129
file wildcards, history lists,
329-330
file-related commands (emacs
editor), 299-302
filenames, 46
absolute vs relative, 51-52
rc suffix, 54
searching with find command,
400-401
sorting alphabetically, 152
suffixes, appending (ls command), 67
wildcards, 162-166
character range, 164-166
echo command, 162
162-164
spaces, 166
files
aliases, 67
archive random library, 132
autoexec.bat (DOS file), 50
awk command, 177
boot, 49
changes, saving (emacs editor),
288
command.com (DOS file), 50
compress program, 83-84
config.sys (DOS file), 50
configuration files (C shell),
317-320
contents, viewing (cat program),
136-139
copying (cp command), 116-117
creating with touch command,
78-79
.cshrc file
319-320
displaying contents, 137
viewing with more
command, 139
deleting text (vi editor), 220-229
dd command, 224-225
dw command, 222-223
u command, 221-222
dynix, 49
/etc/printcap, 380-381
finding specified printers, 383
/etc/shells, 312
execute permission, 90
.exrc file (vi editor), 260-264
262-263
listing abbreviations in
effect, 263
filenames
suffixes, appending (ls
command), 67
wildcards, 162-166
formats, binary, 46
group
109-110
551
header files, 511
hidden, 52-53
inserting text with vi editor,
212-220
append (a) command,
216-218
insert (i) command, 216-217
O command, 215-216
o command, 214-215
Internet files (ftp program),
483-493
483
Apple Computer FTP
archive, 489-490
MIT AI Laboratory
487-488
484-486
starting ftp, 484
large files, viewing (more
program), 139-143
lib.b file, 132
listing
ls command, 65
log, write-only permissions, 89
.login file
317-319
cp command, 117
moving (mv command), 118-119
moving around in
vi editor, 205-211
numbering, awk command (NR
variable), 191
numbering lines, 153-154
cat program, 153-154
nl command, 153-157
vi editor, 255-257
owner
changing (chown command),
108-109
preference, 52
printing, 384-387
552
files
aliases, 386
print queue, 385-386,
391-394
printer information, 386-387
specifying printers, 384-385
removing (rm command),
121-123
precautions, 123-125
removing extraneous lines (uniq
command), 149-150
common lines, 149
duplicate lines, 149-150
renaming (mv command),
119-120
saving (vi editor), 218
searching
find command, see find
command
searching and replacing
(emacs editor), 294-297
vi editor, 229-234
sizes, indicating (ls command),
65-66
blocks occupied, 66
sorting
changing sort order (ls
command), 71-73
sorting information (sort
command), 150-153
stripping, 131
symbolic links, 68
text, deleting, 289-293
text files (awk command),
188-195
unix, 49
viewing
file lines, 133-136
file redirection, 146-147
vmunix, 49
word counts (wc program),
147-149
filters, 158
file information, sorting, 150-153
filenames, sorting alphabetically,
152
find command, 398-403
-exec flag, 402-403
flags, 398
-l flag, 404
listing files/directories, 398-399
-mtime n flag, 400
-name flag, 399-400
findgroup alias (rn program),
469-470
finger command, 446
finger program (hidden files), 55
flags, 40
-?, 14, 16
-1 (ls command), 68, 77, 152
-a, 54
du command, 80-81
ls command, 67-68, 77
ps command, 368
-ag (ps command), 372-373
-atime n (find command), 398
-ax (ps command), 373
-b
sort command, 151
test command, 354
-C (ls command), 68, 77
-c
grep command, 172-174
more program, 139
test command, 354
uniq command, 150
-CF (ls command), 486
combining, 68-69
-ctime n (find command), 398
-d, 93
lp command, 385
ls command, 71, 78
more program, 139
sort command, 151
test command, 354
uniq command, 150
-dptr (lp command), 384
-e (test command), 354
-ef (test command), 354
-eq (test command), 352
-exec (find command), 403
exec command, 398
-F (ls command), 67-68, 78
-f
awk command, 188
pr command, 388-390
sort command, 151
test command, 354
-Fc (awk command), 188
find command, 398
-G (test command), 353
-g
Is command, 107
ps command, 368
test command, 354
g (sed command), 180
-ge (test command), 352
-gt (test command), 352
-h, 14, 16
lpr command, 384
-hhdr (pr command), 388
-i
lpr command, 384
rm command, 122
-k, 7
stest command, 354
-L
lpr command, 384
test command, 353
-l
find command, 404
Is command, 107
kill command, 375
ls command, 74-78
ps command, 369
-l math, 38
-le (test command), 352
-lt (test command), 352
-m
ls command, 68, 78
pr command, 388
-mtime n (find command),
398-400
multi-letter flags (du command),
81
+n (pr command), 388
-n
pr command, 388
sort command, 151-152
n (mesg command), 408-409
-name (find command), 399-400
-name pattern (find command),
398
-ne (test command), 352
-nt (test command), 354
-O (test command), 354
-ot (test command), 354
-P
lpr command, 385
print queues, 393
grep command
-p
mkdir command, 10
test command, 354
-Pn (lp command), 384
-Ppr (lpr command), 384
-print (find command), 398
-R
lpr command, 384
vi editor, 234
-r
ls command, 72, 78
rm command, 122
test command, 354
-S (test command), 354
-s
more program, 139
nl command, 156-157
sending mail (mailx), 417
test command, 354
sort command, 151
-t (ls command), 72, 78
-t xx (ps command), 369
-ttitle (lp command), 384
-type c (find command), 398
-u
ps command, 369, 372
test command, 354
uniq command, 150
-user name (find command), 398
-v
cat program, 136
lpinfo command, 386-387
-w
ps command, 369
test command, 354
-wn (pr command), 388
-x
ps command, 369
test command, 354
y (mesg command), 408
floating-point variables, 513
modifiers, 514
floppy device driver, 47
floppy disks (man page), 10
flow control, 294, 303
folders, 45
mail folders, file command, 129
for command, 358
for loop, 520-521
foreach loop, 343
foreground job, 377
formats, binary, 46
formatting print jobs (pr
command), 387-391
comparing directory contents,
389-390
header information, adding, 389
two-column mode, 388
FORTRAN, 5
forwarding e-mail messages
(mailx), 414
Free Software Foundation (FSF),
308
From: line, basic notations, 445
FTP, 426
FTP archives
Apple Computer FTP archive,
489-490
MIT Artificial Intelligence
Laboratory, 487
pub directory, 487-488
Table 23.2, 491-492
ftp program
anonymous FTP archive,
487-490
Apple Computer FTP
archive, 489-490
Laboratory, 487-488
binary mode (transferring files),
490-491
commands, 483-484
copying Internet files, 483-493
483
dir command, 484-485
flags, 486
get command, 486
ls command, 484-485
flags, 486
mget command, 486
484-486
starting, 484
function libraries, 47
functions (C programming
language), 521-523
macro replacements, 523
553
G
^g command (vi editor), 256
G command (vi editor), 229-231
g command (rn program), 472
-G flag (test command), 353
-g flag
Is command, 107
ps command, 368
sed command, 180
test command, 354
-ge flag (test command), 352
GECOS (GE computer operating
system), 4
get command
ftp program, 484
gid (group ID), 30
GNU EMACS, 283
go-to-line command (vi editor),
229-231
gopher program, 454-460
bookmarks, adding, 456-457
electronic books, accessing,
458-459
electronic mailing of files, 458
gopherspace, 478
introduction screen, 454-455
library searches, 460-466
National Library in
Venezuela, 460-461
Queensland University of
Technology, 461-463
University of California,
464-466
logging on, 454-455
navigating, 455-458
notation, 458
Gopher protocol, 426
-c flag, 172-174
general form, 173
-i flag, 172-173
-l flag, 172-174
-n flag, 172-174
searching
inclusion ranges, 168
punctuation sequences,
170-171
specific words, 168
554
grep program, xargs command
grep program, xargs command,
403-405
groups
directories, identifying, 107-108
files
109-110
group ID, 30
id command, 30
permissions, 98
-gt flag (test command), 352
H
~h command (mailx), 419
h command, 141
vi editor, 219
-h flag, 14-16
lpr command, 384
hang-up signal (kill command),
376
hard disks, mnt and sys
directories, 48
head program
pipelines, 134-135
viewing file lines, 133-135
multiple files, checking, 134
-n format, 134
header files, 511
headers (e-mail messages), 414
headers command (mailx), 414-415
help
bc command, 37
help command, 14-17
man pages, 7-14
Help options (emacs editor), 287
help systems (emacs editor),
297-299
options, 297-298
heuristics, 40
passwd command, 26
Hewlett-Packard, 5
-hhdr flag (pr command), 388
hidden files, 52-53, 61
history commands, 329
C shell, 333
exclamation points, 327
repeating commands, 327
history lists (shells), 327-333
building, 327-328
csh history mechanism, 330-331
echo command, 330
ksh history-edit command mode,
332-333
repeating commands, 328-329
history mechanisms (shells),
324-327
shell parameters, checking status,
325
starting, 325
HoloNet Services Gateway, 500
home directory, 49, 61
HOME variable, 57
cd command, 59
environment variables, 314
host name, 40
HTML (hypertext markup
language), 427
HTTP (hyper-text transfer
protocol), 427
hyperlinks, 440
I
i command (vi editor), 216-219
I command (emacs help system),
298
-i flag
lpr command, 384
rm command, 122
I value (process status values), 371
i-list, 4
i-nodes, 4
id command, 30
if command, 355-356
if statements, 517-519
if-then condition (awk command),
194
inclusion range, 185
indeterminate loops, 357, 360
infinite loops, 520
infix calculators, 36-38
infix notation, 36
Information Highway, 480
information servers list Web
site, 529
initial passwords, 23
input (standard input), 151
insert command (vi editor),
216-219
insert mode (vi editor), 213, 243
insertion commands (vi editor),
219-220
integer variable, 512
modifiers, 513-514
interactive program, 143
interactive shells, 320
interfaces (command-line), 2
Internet, 426-440
copying files (ftp program),
483-493
483
Apple Computer FTP
archive, 489-490
MIT AI Laboratory
487-488
484-486
starting ftp, 484
domain naming, 478
e-mail
addresses, 444-445
d suffix, 447
finger command, 446
From: line, 445
sending to Internet users,
444-446
see also e-mail
bookmarks, adding, 456-457
electronic books, accessing,
458-459
electronic mailing of
files, 458
introduction screen, 454-455
library searches, 460-466
logging on, 454-455
navigating, 455-458
notation, 458
Korn shell
Netscape Navigator
Bookmark menu, 431
customizing, 437-440
local files, specifying
(command line), 430-431
starting, 427-431
URLs, specifying (command
line), 429-430
What’s New? button, 429
newsgroups, see newsgroups
protocols, 426-427
remote sites (connecting to),
480-483
logging in to, 481
logging out, 481
rlogin command, 480
telnet commands, 481-483
users, talk system, 446-449
Scott Yanoff’s List of Internet
Services, 506
searching Web sites, 432-437
AltaVista (Web crawler),
436-437
Yahoo! (search engine),
432-435
telnet sites, 499-506
Book Stacks Unlimited,
503-506
Compact Disk Connection,
500-503
Usenet, 466-467
database of newsgroups (C
shell alias), 467
WAIS (Wide Area Information
Services), 449
accessing, 449-451
Web sites, see Web sites
Center, 496
logging in, 496-497
whatis command (mail searches),
497-499
Internet Relay Chat (IRC)
protocol, 426
Internet Services List Web site, 529
Is command
-g flag, 107-108
-l flag, 107
555
J
K
j command (vi editor), 219
job control, 321
command shells, 306-307
job ID, 365
jobs, 377
print jobs, formatting (pr
command), 387-391
389-390
header information,
adding, 389
running in background, 365-368
& symbol (moving
automatically), 367
automatic starts, 366
input/output, 368
job ID, 365
z command, 367
stopping, 362-365
fg command, 362-364
363-364
terminating (kill command),
374-377
376
logging out, 377
specifying signals to kill
command, 374-375
tracking (ps command), 368-374
-ag flag, 372-373
-ax flag, 373
BSD-style flags, 368-369
status values, 371
-u flag, 372
zombie process, 371
see also processes
jobs command, 369
jsh (command shell), 307
Julian calendar, 34
k command (vi editor), 219
K command (emacs help system),
298
-k flag, 7
test command, 354
kernel, 61
Kernighan, Brian, 3
key bindings, 303
keys
b key (moving around files word
by word), 211
control-key commands (vi
editors), 208, 209
Control-f command,
209-210
Control-u command, 210
cursor control keys (vi editor),
205-208
mapping, 279
Clear key, 264
.exrc file, 260-264
w key (moving around files word
by word), 211
keyword searches (-k flag), 7
kill command, 370, 374-377
repeating, 376
signals, 374-375
376
Korn shell, 2, 307
$ prompt, 24
aliases, 335-337
!* notation, 340
cd command, 337
chdir command, 337
connecting to accounts on
other systems, 336-337
command-alias mechanism,
333-335
DOS commands,
re-creating, 334
556
Korn shell
general format, 333
history list, 327-333
building, 327-328
csh history mechanism,
330-331
echo command, 330
ksh history-edit command
mode, 332-333
repeating commands,
328-329
history mechanism, 324-327
shell parameters, checking
status, 325
starting, 325
jobs (processes)
fg command, 362-364
running in background,
365-368
363-364
stopping, 362-365
cd alias, 339-340
cwd variable, 338
values, 338
ksh (command shell), 307
ksh history-edit command mode
(Korn shell), 332-333
L
-l command (archie program), 495
l command (vi editor), 219
L command (emacs help system),
298
-L flag
lpr command, 384
test command, 353
-l flag
find command, 404
Is command, 107
kill command, 375
ls command, 74-78
ps command, 369
-l math flag, 38
lcd dir command (ftp program),
484
-le flag (test command), 352
left rooted patterns, 185
lib.b file, 132
libraries
math, 47
searching with gopher program,
460-466
National Library in
Venezuela, 460-461
Technology, 461-463
464-466
line numbering, 153-154
nl command, 153-157
option, 156
-s flag, 156-157
vi editor, 255-257
listing
directories (ls command), 69-71
directory trees, listing recursively,
73-74
dot files, 67
files, 65
local printers, finding with printer
script, 380-383
log files, write-only permissions, 89
logging in, 22
remote systems
rlogin command, 480-481
logging out
background processes, 377
exit command, 24
remote systems
rlogin command, 481
telnet command, 482
.login file (C shell configuration
files), 317-319
login shells, 321, 372, 378
awk command, 189
choosing new shells, 310-313
available shells, identifying,
310-311
change shell command,
312-313
valid shells, confirming, 312
visual shell, 311
identifying system’s shell,
309-310
viewing default login shells,
308-309
LOGNAME variable (environment
variables), 316
long directory listings (ls
command), 75-78
long integers (C programming
language), 513
long listing format (ls command)
directories, 75-78
files, 74
looping expressions, 357-359
for command, 358
while loop, 358-359
looping statements, 520-521
do loop, 521
for loop, 520-521
while loop, 520
loops, 360
foreach loop, 343
lp command
flags, 384
aliases, 386
print queue, 385-386
lpinfo command, 386-387
lpq command (print queues), 391
lpr command
flags, 384
aliases, 386
print queue, 385-386
lprm command (print queues), 391
ls -C -F command, 60
ls -C -F / command, 46
./ls command, 56
ls command, 53, 64-65
-1 flag, 68, 77
-a flag, 68, 77
-g flag, 107
-C flag, 68, 77
combining flags, 68-69
-d flag, 78
listing directories, 71
modifiers
directories, listing, 69-71
directory trees, listing recursively,
73-74
-F flag, 68, 78
filename suffixes,
appending, 67
file type indicators, 89
files
listing, 65
sizes, indicating, 65-66
flags, 77-78
ftp program, 484
484-486
-l flag, 74-78
long listing format
directories, 75-78
files, 74
-m flag, 68, 78
mkdir command, 118
modifying output with awk
command, 192
numeric permissions strings, 103
permissions, changing, 97
-R flag, 78
directory trees, listing
recursively, 73-74
-r flag, 72, 78
rm command, 120
-s flag, 68, 78
blocks occupied by files, 66
file size, indicating, 65-66
sort order, changing, 71-73
-t flag, 72, 78
-x flag, 71-72, 78
-lt flag (test command), 352
Lukasiewicz, Jan, 36
M
M command (emacs help
system), 298
-m flag
ls command, 68, 78
pr command, 388
M-< command (emacs editor), 289
M-> command (emacs editor),
288-289
M-a command (emacs editor), 289
M-b command (emacs editor), 289
M-d command (emacs editor), 293
M-Delete command (emacs editor),
293
M-e command (emacs editor),
286, 289
M-f command (emacs editor), 289
M-k command (emacs editor), 293
M-v command (emacs editor), 289
Macintosh
directory delineator, 51
file system, 50
hidden files, 53
moving files, 116
Preferences folder, 52
System Folder, 45
Macintosh interface, 2
macro replacements (C programming language), 523
mail, 411
mail folders, 424
file command, 129
mail headers, 424
MAIL variable (environment
variables), 316
mailbox, 424
mailx
headers, 414
quit command, 416
reading e-mail, 411-416
save command, 415
~h command, 419
~p command, 418
~r command, 418
major number (device drivers), 144
man pages, 7-14
apropos command, 12
fdformat command, 12
mkdir command, 9
more program, 14
sort command, 13
System V organization, 8
557
map command (vi editor)
mapping keys (vi editor)
arrow keys, 265
Clear key, 264
.exrc file, 260-264
262-263
listing abbreviations in
effect, 263
math functions (sine), 38
math library, 47
mathematical functions (variables),
350-351
mathlw queue, 391-392
McIlroy, 4
mesg command (enabling
messages), 408-409
mesg y variable, 319
messages, remove login.copy?, 119
Meta key, 282, 303
mget command
ftp program, 484
specifying wildcard patterns, 486
MH shell, 307
MIME (multimedia Internet mail
extension), 427
minor number (device drivers), 144
MIT, 3
Laboratory (anonymous FTP
archive), 487
pub directory, 487-488
mkdir command, 114-116
man page, 9
-p flag, 10
mnt directory, 48
modal editors, 201
modal programs, 243
modeless editors, 201
modeless programs, 244
modem protocols, 467
modes (vi editor), 201
modifiers
character variable, 514
floating-point variable, 514
integer variable, 513-514
558
more program
more program
entering commands, 142
man command, 14
viewing large files, 139-143
-More- prompt commands, 141
Motif, 5
motion commands
emacs editor, 289
vi editor, 219-220
moving
files (mv command), 118-119
sentences and paragraphs (vi
editor), 266-270
( command, 268
) command, 268
/ command, 268
{ command, 270
mput command (ftp program), 484
MS-DOS interface, 2
msh (command shell), 307
-mtime n flag (find command),
398-400
multi-letter flags (du
command), 81
multichoice system, 2
Multics, 3
multitasking, 2
multiuser systems, 2, 5-6
mv command, 116
directories, renaming, 120
files
moving, 118-119
renaming, 119-120
N
\n character (C programming
language), 511
:n command (vi editor), 236, 240
n command, 141
emacs editor, 296
vi editor, 232-234
+n flag (pr command), 388
-n flag
cat program (numbering file
lines), 153-154
pr command, 388
n flag (mesg command), 408-409
-n format (head command), 134
n[Return] command, 141
-name flag (find command),
399-400
-name pattern flag (find command), 398
NAME variable (environment
variables), 316
named emacs command, 294, 303
naming directories, 115
National Library in Venezuela, 460
navigating
file system, 58
Navigator (Netscape)
Bookmark menu, 431
customizing (General Preferences), 437-440
Appearance tab, 437
Fonts tab, 437
Helpers tab, 439
Images tab, 439
starting, 427-431
troubleshooting, 428
line), 429-430
X Window System, 427
-ne flag (test command), 352
net directory, 49
Netnews, 426
Netscape Navigator
Bookmark menu, 431
customizing, 437-440
customizing (General Preferences)
Appearance tab, 437
Fonts tab, 437
Helpers tab, 439
Images tab, 439
starting, 427-431
troubleshooting, 428
line), 429-430
X Window System, 427
networks (multiusers), 29
newmail variable, 319
newsgroups, 467, 478
alt newsgroups, 467
hierarchies, 467
rn program, 468
findgroup alias, 469-470
RNINIT environment
variable, 469
start up, 470-471
starting options, 468
SUBJLINE environment
variable, 474
rn program (newsgroups)
= command, 474
g command, 472
reading articles, 472-475
unsubscribing from
newsgroups, 471-472
tin program, 475-476
next search command (vi editor),
232-234
next-file command (vi editor), 240
nf command, 141
NF variable (awk command),
190-191
nl command
option, 156
-s flag, 156-157
noninteractive shells, 320
Norton Utilities (rescuing DOS
files), 123
notations
!* notation (aliases), 340
+= notation, 193
egrep command, 175
gopher program, 458
math notations, 36
RPN, 36
NR variable (awk command), 191
ns command, 141
-nt flag (test command), 354
null characters, 144
numbering files
awk command (NR
variable), 191
numbering lines (files), 153-154
nl command, 153-157
vi editor, 255-257
PRINTER environment variable
numbering systems, 38
numbers, computing sums, 193
numeric comparisons (test
command), 352
numeric equivalents (permissions),
100
numeric permissions strings,
calculating, 102-104
numeric repeat prefixes (vi editor),
253-255
O
O command (vi editor),
215-216, 219
o command (vi editor),
214-215, 219
-O flag (test command), 354
\ooo character (C programming
language), 511
open command (ftp program), 484
Open Desktop, 5
OpenWindows, 5
operating systems (CTSS), 4
operators (C programming
language)
binary operators, 515
unary operators, 514-515
-ot flag (test command), 354
output
redirecting, 146-147
standard output, 151
output formats (Sun Microsystems
workstation), 131-132
P
p (dc command), 38
~p command (mailx), 418
-P flag
lpr command, 385
print queues, 393
-p flag
mkdir command, 10
test command, 354
paragraphs
assigning to UNIX commands (vi
editor), 272-273
moving (vi editor), 266-270
( command, 268
) command, 268
/ command, 268
{ command, 270
tightening lines (fmt command),
273-274
passwd command, 25
password entry, 124-125
passwords, 6
changing (passwd command), 25
choosing, 26-27
initial, 23
PATH variable, 57
environment variables, 315
pathnames, absence of, 4
/pattern command, 141
PC
directory delineator, 51
file system, 50
pcfs directory, 49
PDP-7, 4
PDP-11, 4
percent sign (%), 24
permission strings, 74-75, 85
permissions, 92-93
absolute numeric values, 98
binary convention, 99
binary equivalents, 100
concentric circles of access, 98
default (umask command), 104,
106-107
deleting directories, 120-121
directory, 93-96
execute, 93-94
execute-only, 94
read, 93-94
write-only, 95
file, 91
files, 88-90
execute, 90-91
read, 89
write, 89
96-98
numeric equivalents, 100-103
numeric strings, calculating,
102-104
parts, 88
setting new (chmod command),
98-102
pipe (|) character, 134-135
559
pipelines, 134-135, 144
cat program, 138
re-routing with tee command,
196-197
plus sign (vi editor), 234
-Pn flag (lp command), 384
pointers (C programming
language), 524-526
postfix calculators, 38-39
postfix notation, 36
-Ppr flag (lpr command), 384
pr command
-f flag, 390
flags, 388
formatting print jobs, 387-391
389-390
header information,
adding, 389
piping output to lpr
command, 390-391
preference files, 52, 61
prefixes, adding/removing with sed
command, 184
-print flag (find command), 398
print job name, 395
print jobs
formatting (pr command),
387-391
389-390
header information,
adding, 389
removing from print queues, 393
print queue, 385-386, 391-394
limiting output, 392-393
print requests
removing, 394
resubmitting, 393-394
printer status, checking, 393
removing print jobs, 393
printenv command, 57
printer configuration files, /etc/
printcap file, 380-381
PRINTER environment variable,
383-385
560
printers script
printers script (finding local
printers), 380-383
awk script, creating, 381-382
PRINTER environment
variable, 383
rehash command, 382
printf function (C programming
language), 511
aliases, 386
print queue, 385-386, 391-394
procedural libraries, 47
processes, 378
errant processes, 377
running in background, 365-368
& symbol (moving
automatically), 367
automatic starts, 366
input/output, 368
job ID, 365
z command, 367
stopping, 362-365
fg command, 362-364
363-364
terminating (kill command),
374-377
376
logging out, 377
command, 374-375
tracking (ps command), 368-374
-ag flag, 372-373
-ax flag, 373
BSD-style flags, 368-369
status values, 371
-u flag, 372
zombie process, 371
wedged processes, 378
zombies, 378
see also jobs
programming languages, 5
programs
archie, see archie program
cat
-n flag, 153-154
pipelines, 138
-s flag, 136-137
-v flag, 136
viewing file contents,
136-139
ftp, see ftp program
head
-n format, 134
pipelines, 134-135
viewing file lines, 133-135
more
entering commands, 142
viewing large files, 139-143
tail, viewing file lines, 135-136
vi editor, 200-205
colon commands, 236-241
control-key commands,
208-209
cursor control keys, 205-208
deleting text, 220-229
insert mode, 213
modes, 201
motion and insertion
commands, 219-220
read-only format, 234-235
saving files, 218
searching files, 229-234
starting, 202-203
startup options, 234-236
TERM environment variable,
setting, 203-204
text, inserting into files,
212-220
transposing case
(~command), 228
unknown-terminal-type error
message, 203
viewing file contents,
204-205
wq command, 229
wc, 147-149
see also commands
prompt command (ftp program),
484
prompts
$, 24
cd alias, 339-340
cwd variable, 338
values, 338
protocols, 467, 478
Internet, 426-427
ps command, 368-374
-ag flag, 372-373
-ax flag, 373
flags (BSD-style), 368-369
Sequent workstation output, 374
Sun workstation output, 374
-u flag, 372
punctuation in commands, 28
punctuation sequences, searching,
170-171
put command (ftp program), 484
pwd command, 58
ftp program, 484
Q
:q command (vi editor), 220, 237
q command, 141
emacs editor, 296
:q! command (vi editor), 220, 237
Technology, 461
query-replace feature (emacs
editor), 294-296
options, 296
question mark (filename
wildcards), 162-164
echo command, 162
162-164
queuing systems (printing),
391-394
limiting output, 392-393
print requests
removing, 394
resubmitting, 393-394
printer status, checking, 393
removing print jobs, 393
quit command (mailx), 416
quiting emacs editor, 283
quotation marks (nl
command), 157
quoted text, 478
search path
R
\r character (C programming
language), 511
:r command (vi editor), 238-239
~r command (mailx), 418
R command (replace commands),
246
r command (replace commands),
246
characters, changing, 248-249
:r filename command (vi editor),
237
-R flag
lpr command, 384
ls command, 78
directory trees, listing
recursively, 73-74
vi editor, 234
-r flag
ls command, 72, 78
rm command, 122
test command, 354
R value (process status values), 371
rc suffix, 54
re-routing pipelines (tee command), 196-197
read command, assigning
variables, 349
read Netnews (rn) program, 317
read permissions
directory, 93-94
file, 89
read-file command (vi editor),
238-239
read-only format (vi editor),
234-235
reading e-mail
Elm Mail System, 421
mailx, 411-416
headers, 414
quit command, 416
save command, 415-416
recipes, databases, 452-453
recursive command, 125
redirecting files, 146-147
regular expressions, 167, 185
notations, 167
sed command, 182-183
rehash command, 91
printers script, 382
relative filenames, 51-52, 61
remote Internet sites
connections, 480-483
logging in, 481
logging out, 481
rlogin command, 480
users, learning information,
447-448
remote system
d suffix, 447
Internet, talk system, 446-449
users, talking with, 448
removable cartridge drives, 48
remove login.copy? message, 119
removing, see deleting
renaming
directories (mv command), 120
files (mv command), 119-120
repeating commands (vi editor),
253-255
replace command, 246-252
characters, changing, 248-249
replace mode (vi editor), 279
Request for Comment, 507
responding to e-mail
mailx, 414
restoring deleted text (emacs
editor), 293
Return command (vi editor), 219
Return key (vi editor), 201
Ritchie, Dennis, 3
rlogin command, 480-481
rm command, 121-123
precautions with, 123-125
rmdir command, 120-121
rn program (newsgroups), 317, 468
findgroup alias, 469-470
g command, 472
reading articles, 472-475
= command, 474
RNINIT environment
variable, 469
start up, 470-471
starting options, 468
561
SUBJLINE environment
variable, 474
unsubscribing from newsgroups,
471-472
RNINIT environment variable (rn
program), 469
root directory, 45, 62
row-first order, 85
RPN (reverse Polish notation), 36
rsh (command shell), 307
S
-s command (archie program), 495
S command (emacs help system),
298
-S flag (test command), 354
-s flag
ls command, 68, 78
blocks occupied by files, 66
file size, indicating, 65-66
more program, 139
nl command, 156-157
sending mail (mailx), 417
test command, 354
S value (process status values), 371
save command (mailx), 415
saving
file changes (emacs editor), 288
files (vi editor), 218
key mappings, 265
SCO, 2
Scott Yanoff’s List of Internet
Services, 506
stopping/starting jobs, 363-364
scripts (shell scripts), 340-344
creating, 342-343
printers
finding local printers,
380-383
PRINTER environment
variable, 383
rehash command, 382
scrolling man pages (alias
command), 14
search path, 62
562
search-and-replace command
search-and-replace command (vi
editor), 257-260
changing
all occurrences within files,
259-260
words, 258-259
searches
databases, searching with WAIS,
449-454
directories, 56
egrep command, 175-176
notation, 175
expanding file searches, 170
fgrep command
awk command, 177
178-179
multiple patterns, 176-179
character range, 164-166
echo command, 162
162-164
spaces, 166
find command, 398-403
-exec flag, 402-403
flags, 398
-l flag, 404
listing files/directories,
398-399
-mtime n flag, 400
-name flag, 399-400
-c flag, 174
general form, 173-174
-i flag, 173
inclusion range searches, 168
-l flag, 174
-n flag, 174
punctuation sequences,
170-171
specific word searches, 168
library searches (gopher
program), 460-466
National Library in
Venezuela, 460-461
Technology, 461-463
464-466
PATH directories (shell scripts),
343-344
PATH variable, 90
searching and replacing (emacs
editor), 294-297
294-296
vi editor, 229-234
? command, 232-233
/ search command, 231-232
G command, 229-231
n command, 232-234
Web sites, 432-437
AltaVista (Web crawler),
436-437
Yahoo! (search engine),
432-435
searching and replacing
294-296
security
passwords, 26-27
permission strings, 74-75
sed command, 179-184
assigning paragraphs to UNIX
commands (vi editor), 272-273
deleting lines in streams, 181-182
g flag, 180
prefixes, adding/removing, 184
command, 181
substituting patterns, 180
modifying names, 180
separating, 181
sending e-mail (mailx), 417-420
~h command, 419
~p command, 418
~r command (reading in
files), 418
sendmail program, 132
sentences, moving (vi editor),
266-270
( command, 268
) command, 268
/ command, 268
{ command, 270
separator flag (nl command),
156-157
Sequent computers, 131
Sequent workstation, ps command
output, 374
set commands (configuration
options), 319
:set number command (vi editor),
256
setprompt alias (customizing
prompts), 339-340
setprompt command, aliases, 337
shell alias, 126
restoring deleted files, 123
shell scripts, 92, 279, 340-344
creating, 342-343
printers
finding local printers,
380-383
PRINTER environment
variable, 383
rehash command, 382
SHELL variable (environment
variables), 314
shells, 6, 306-309
Bourne shell, 306-307
Bourne Again shell, 307
jsh shell, 307
C shell, see C shell
case command (conditional
expressions), 357
changing (chsh command), 124
choosing new shells, 310-313
available shells, identifying,
310-311
change shell (chsh)
command, 312-313
valid shells, confirming, 312
visual shell (vsh), 311
command aliases, 306-307
command history, 306-307
default login shells, viewing,
308-309
telnet command
environment, 313-317
variables, 314-317
expanding file searches, 170
identifying system’s shell,
309-310
if command (conditional
expressions), 355-356
interactive shells, 320
job control, 306-307
jobs, see jobs
Korn shell, see Korn shell
login shells, 372
looping expressions, 357-359
for command, 358
while loop, 358-359
noninteractive shells, 320
test command (comparison
functions), 351-355
file system status tests,
353-355
test operators, 352
353-354
variables, 348-350
arithmetic functions (expr
command), 350-351
assigning, 349-350
setting values, 348-349
visual (vsh) shell, 311
short integers (C programming
language), 513
signals, 378
catching, 376
signed modifier (C programming
language), 513
sine (bc command), 38
single quote (‘), 124
slash (/), 62
filenames, 46
slash directory, 45, 62
software description database
(archie program), 496
Solaris, 2
sort command
filenames, sorting alphabetically,
152
flags, 151
man page, 13
-n flag, 152
-r flag, 153
sorting file information, 150-153
sorting
data (awk command), 190
files, changing sort order (ls
command), 71-73
[Space] command, 141
spacing in commands, 28
spell command, 178-179
standard error, 151, 158
standard input, 151, 158
standard output, 151, 158
starting emacs editor, 282-284
starting flag, 424
startup options (vi editor), 234-236
plus sign (+), 234
-R flag, 234
statements
break; (exiting loops), 521
conditional, 517-520
continue; (restarting loops), 521
if, 517-519
looping, 520-521
do loop, 521
for loop, 520-521
while loop, 520
switch, 519-520
stream editor (sed command),
179-184
deleting lines in streams, 181-182
g flag, 180
separating, 181
prefixes, adding/removing, 184
command, 181
substitute command, 180
streams, deleting lines with sed
command, 181-182
string comparisons (test command), 353
stripping files, 131
structures (C programming
language), 526-528
union structures, 528
stty commands (configuration
options), 318
stty tostop command, 365
subdirectories, beginning of
UNIX, 4
SUBJLINE environment variable
(rn program), 474
563
subshell, 321
substitute command, 179-180
Sun Microsystems, 5
workstation, output formats,
131-132
Sun system mail folders, 129
Sun workstation (ps command
output), 374
surfing, 440
switch statements, 519-520
Symantec Utilities (Macintosh),
123
symbolic links, 46, 62, 68, 131
symbolic notation (chmod
command), 96
SYMMETRY i386, 131
sys directory, 48
system administrator, initial
passwords, 23
system prompts, setting custom
prompts, 338-340
cd alias, 339-340
cwd variable, 338
values, 338
System V
man page organization, 8
System V Release 4, 2
T
\t character (C programming
language), 511
T command (emacs help system),
298
-t flag (ls command), 72, 78
T value (process status values), 371
-t xx flag (ps command), 369
tables, building with awk
command, 194-195
tail program, viewing file lines,
135-136
talk system (Internet), 446-449
tcsh (command shell), 307
tee command, re-routing pipelines,
196-197
Telnet, 426
telnet command, 480
connecting to remote systems,
481-482
archie system, 496
logging out, 482
options, 482
564
telnet sites
telnet sites, 499-506
Book Stacks Unlimited, 503-506
Compact Disk Connection,
500-506
TERM environment variable,
314-315
setting (vi editor), 203-204
terminating processes (kill
command), 374-377
logging out, 377
command, 374-375
376
test command, 351-355
file system status tests, 353-355
test operators, 352
353-354
text
change command, 246-252
247-248
249-252
deleting
deleting (vi editor), 220-229
dd command, 224-225
dw command, 222-223
u command, 221-222
inserting into files (vi editor),
212-220
append (a) command,
216-218
insert (i) command, 216-217
O command, 215-216
o command, 214-215
moving sentences and paragraphs
( command, 268
) command, 268
/ command, 268
{ command, 270
paragraphs, see paragraphs
replace command, 246-252
248-249
restoring deleted text (emacs
editor), 293
transposing case (vi editor), 228
text files
awk command, 188-195
awk program files,
creating, 194
BEGIN pattern, 192
login shells, 189
modifying output of ls
command, 192
NR variable, 191
sorting data, 190
numbering, awk command
(NR variable), 191
tftpboot directory, 49
Thompson file system, 4
Thompson, Ken, 3
tilde commands
mailx, 417-418
vi editor, 228
time
24-hour time, 31
displaying, 33
time command, 33
tin program (newsgroups),
475-476
TMG programming language, 5
tmp directory, 48
top-level Internet domains, 444
touch command, 91
files, creating, 78-79
permissions, changing, 97
transposing case (vi editor), 228
troubleshooting
starting Netscape Navigator, 428
-ttitle flag (lp command), 384
tty, 29
tty command, 408-409
tty devices, 408-409
-type c flag (find command), 398
-type specifier (find command),
401-402
typos, editing (vi editor), 225-227
tilde (~) command, 228
U
û command (vi editor), 220
u command (vi editor),
221-222, 252
-u flag
ps command, 369, 372
test command, 354
uniq command, 150
U.S. State

Teach Yourself UNIX

Transcription

Similar documents

ordering instructions mail order

home on the dunes - Gulfview Construction

Koperasi Kastam Diraja Malaysia Berhad (KOSKAM)

EurekaWeb - Eureka Support

Preview - stanfordhouse.com.hk

AIRMAN`S INFORMATiON MANUAL

We`ll help you get rid of junk mail! Please place unwanted catalogs

Far Ahead of his Time

jason mraz - Conscious Connection

FLF For Sale - FLF holdings